Si Reference

LabVIEW
TM
System Identification Toolkit Algorithm References

System Identification Toolkit User Manual
June 2008
Support
Worldwide Technical Support and Product Information
ni.com
National Instruments Corporate Headquarters
11500 North Mopac Expressway Austin, Texas 78759-3504 USA Tel: 512 683 0100
Worldwide Offices
Australia 1800 300 800, Austria 43 662 457990-0, Belgium 32 (0) 2 757 0020, Brazil 55 11 3262 3599,
Canada 800 433 3488, China 86 21 5050 9800, Czech Republic 420 224 235 774, Denmark 45 45 76 26 00,
Finland 358 (0) 9 725 72511, France 01 57 66 24 24, Germany 49 89 7413130, India 91 80 41190000,
Israel 972 3 6393737, Italy 39 02 41309227, Japan 0120-527196, Korea 82 02 3451 3400,
Lebanon 961 (0) 1 33 28 28, Malaysia 1800 887710, Mexico 01 800 010 0793, Netherlands 31 (0) 348 433 466,
New Zealand 0800 553 322, Norway 47 (0) 66 90 76 60, Poland 48 22 3390150, Portugal 351 210 311 210,
Russia 7 495 783 6851, Singapore 1800 226 5886, Slovenia 386 3 425 42 00, South Africa 27 0 11 805 8197,
Spain 34 91 640 0085, Sweden 46 (0) 8 587 895 00, Switzerland 41 56 2005151, Taiwan 886 02 2377 2222,
Thailand 662 278 6777, Turkey 90 212 279 3031, United Kingdom 44 (0) 1635 523545
For further support information, refer to the Technical Support and Professional Services appendix. To comment
on National Instruments documentation, refer to the National Instruments Web site at ni.com/info and enter
the info code feedback.
20042008 National Instruments Corporation. All rights reserved.
Important Information
Warranty
The media on which you receive National Instruments software are warranted not to fail to execute programming instructions, due to defects
in materials and workmanship, for a period of 90 days from date of shipment, as evidenced by receipts or other documentation. National
Instruments will, at its option, repair or replace software media that do not execute programming instructions if National Instruments receives
notice of such defects during the warranty period. National Instruments does not warrant that the operation of the software shall be
uninterrupted or error free.
A Return Material Authorization (RMA) number must be obtained from the factory and clearly marked on the outside of the package before
any equipment will be accepted for warranty work. National Instruments will pay the shipping costs of returning to the owner parts which are
covered by warranty.
National Instruments believes that the information in this document is accurate. The document has been carefully reviewed for technical
accuracy. In the event that technical or typographical errors exist, National Instruments reserves the right to make changes to subsequent
editions of this document without prior notice to holders of this edition. The reader should consult National Instruments if errors are suspected.
In no event shall National Instruments be liable for any damages arising out of or related to this document or the information contained in it.
EXCEPT AS SPECIFIED HEREIN, NATIONAL INSTRUMENTS MAKES NO WARRANTIES, EXPRESS OR IMPLIED, AND SPECIFICALLY DISCLAIMS ANY WARRANTY OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. C USTOMERS RIGHT TO RECOVER DAMAGES CAUSED BY FAULT OR NEGLIGENCE ON THE PART OF
NATIONAL INSTRUMENTS SHALL BE LIMITED TO THE AMOUNT THERETOFORE PAID BY THE CUSTOMER. NATIONAL INSTRUMENTS WILL NOT BE LIABLE FOR
DAMAGES RESULTING FROM LOSS OF DATA, PROFITS, USE OF PRODUCTS, OR INCIDENTAL OR CONSEQUENTIAL DAMAGES, EVEN IF ADVISED OF THE POSSIBILITY
THEREOF. This limitation of the liability of National Instruments will apply regardless of the form of action, whether in contract or tort, including
negligence. Any action against National Instruments must be brought within one year after the cause of action accrues. National Instruments
shall not be liable for any delay in performance due to causes beyond its reasonable control. The warranty provided herein does not cover
damages, defects, malfunctions, or service failures caused by owners failure to follow the National Instruments installation, operation, or
maintenance instructions; owners modification of the product; owners abuse, misuse, or negligent acts; and power failure or surges, fire,
flood, accident, actions of third parties, or other events outside reasonable control.
Copyright
Under the copyright laws, this publication may not be reproduced or transmitted in any form, electronic or mechanical, including photocopying,
recording, storing in an information retrieval system, or translating, in whole or in part, without the prior written consent of National
Instruments Corporation.
National Instruments respects the intellectual property of others, and we ask our users to do the same. NI software is protected by copyright and other
intellectual property laws. Where NI software may be used to reproduce software or other materials belonging to others, you may use NI software only
to reproduce materials that you may reproduce in accordance with the terms of any applicable license or other legal restriction.
Trademarks
National Instruments, NI, ni.com, and LabVIEW are trademarks of National Instruments Corporation. Refer to the Terms of Use section
on ni.com/legal for more information about National Instruments trademarks.
Other product and company names mentioned herein are trademarks or trade names of their respective companies.
Members of the National Instruments Alliance Partner Program are business entities independent from National Instruments and have no agency,
partnership, or joint-venture relationship with National Instruments.
Patents
For patents covering National Instruments products, refer to the appropriate location: HelpPatents in your software, the patents.txt file
on your CD, or ni.com/patents.
WARNING REGARDING USE OF NATIONAL INSTRUMENTS PRODUCTS

(1) NATIONAL INSTRUMENTS PRODUCTS ARE NOT DESIGNED WITH COMPONENTS AND TESTING FOR A LEVEL OF
RELIABILITY SUITABLE FOR USE IN OR IN CONNECTION WITH SURGICAL IMPLANTS OR AS CRITICAL COMPONENTS IN
ANY LIFE SUPPORT SYSTEMS WHOSE FAILURE TO PERFORM CAN REASONABLY BE EXPECTED TO CAUSE SIGNIFICANT
INJURY TO A HUMAN.
(2) IN ANY APPLICATION, INCLUDING THE ABOVE, RELIABILITY OF OPERATION OF THE SOFTWARE PRODUCTS CAN BE
IMPAIRED BY ADVERSE FACTORS, INCLUDING BUT NOT LIMITED TO FLUCTUATIONS IN ELECTRICAL POWER SUPPLY,
COMPUTER HARDWARE MALFUNCTIONS, COMPUTER OPERATING SYSTEM SOFTWARE FITNESS, FITNESS OF COMPILERS
AND DEVELOPMENT SOFTWARE USED TO DEVELOP AN APPLICATION, INSTALLATION ERRORS, SOFTWARE AND
HARDWARE COMPATIBILITY PROBLEMS, MALFUNCTIONS OR FAILURES OF ELECTRONIC MONITORING OR CONTROL
DEVICES, TRANSIENT FAILURES OF ELECTRONIC SYSTEMS (HARDWARE AND/OR SOFTWARE), UNANTICIPATED USES OR
MISUSES, OR ERRORS ON THE PART OF THE USER OR APPLICATIONS DESIGNER (ADVERSE FACTORS SUCH AS THESE ARE
HEREAFTER COLLECTIVELY TERMED SYSTEM FAILURES). ANY APPLICATION WHERE A SYSTEM FAILURE WOULD
CREATE A RISK OF HARM TO PROPERTY OR PERSONS (INCLUDING THE RISK OF BODILY INJURY AND DEATH) SHOULD
NOT BE RELIANT SOLELY UPON ONE FORM OF ELECTRONIC SYSTEM DUE TO THE RISK OF SYSTEM FAILURE. TO AVOID
DAMAGE, INJURY, OR DEATH, THE USER OR APPLICATION DESIGNER MUST TAKE REASONABLY PRUDENT STEPS TO
PROTECT AGAINST SYSTEM FAILURES, INCLUDING BUT NOT LIMITED TO BACK-UP OR SHUT DOWN MECHANISMS.
BECAUSE EACH END-USER SYSTEM IS CUSTOMIZED AND DIFFERS FROM NATIONAL INSTRUMENTS' TESTING
PLATFORMS AND BECAUSE A USER OR APPLICATION DESIGNER MAY USE NATIONAL INSTRUMENTS PRODUCTS IN
COMBINATION WITH OTHER PRODUCTS IN A MANNER NOT EVALUATED OR CONTEMPLATED BY NATIONAL
INSTRUMENTS, THE USER OR APPLICATION DESIGNER IS ULTIMATELY RESPONSIBLE FOR VERIFYING AND VALIDATING
THE SUITABILITY OF NATIONAL INSTRUMENTS PRODUCTS WHENEVER NATIONAL INSTRUMENTS PRODUCTS ARE
INCORPORATED IN A SYSTEM OR APPLICATION, INCLUDING, WITHOUT LIMITATION, THE APPROPRIATE DESIGN,
PROCESS AND SAFETY LEVEL OF SUCH SYSTEM OR APPLICATION.
Algorithm References
This document describes the VIs used to implement the algorithms in the LabVIEW System
Identification Toolkit.
Refer to the LabVIEW Help, available in LabVIEW by selecting HelpSearch the LabVIEW Help,
for more information about the main steps in the system identification process and about how to use the
System Identification Toolkit to create applications that can accomplish the various tasks in the process.
Contents
System Identification Palettes, VIs, and SubVIs ...................................................................................... 2
si_Piecewise Linear Fitting.vi .................................................................................................................. 3
SI Detect Feedback (Array).vi .................................................................................................................. 5
si_AR (FB, LS, YW and PC).vi ............................................................................................................... 6
si_AR Burg Method.vi ............................................................................................................................. 8
si_Est ARX Model (SISO).vi ................................................................................................................. 10
si_Est ARX Model (MISO).vi ................................................................................................................ 12
si_Est ARX Model (MIMO).vi .............................................................................................................. 13
si_Est ARMAX Model (SISO).vi ........................................................................................................... 14
si_Est ARMAX Model (MISO).vi ......................................................................................................... 16
si_Est OE Model (SISO).vi .................................................................................................................... 17
si_Est OE Model (MISO).vi ................................................................................................................... 19
si_Est BJ Model (SISO).vi ..................................................................................................................... 21
si_Est BJ Model (MISO).vi .................................................................................................................... 23
si_Est GL Model (SISO).vi .................................................................................................................... 26
si_Est GL Model (MISO).vi ................................................................................................................... 28
SI Estimate Continuous Transfer Function Model (SISO Array).vi ....................................................... 31
SI Estimate Discrete Transfer Function Model (SISO Array).vi ............................................................ 33
SI Estimate User-Defined Model (MIMO Array).vi .............................................................................. 34
si_Impulse Realization (MIMO).vi ........................................................................................................ 35
Subspace Identification Method via Principal Component Analysis (SIMPCA) ................................... 37
si_Correlation Analysis.vi ...................................................................................................................... 41
SI Estimate Impulse Response (Least Squares Array).vi ....................................................................... 43
si_Frequency Analysis.vi ....................................................................................................................... 45
SI Estimate Partially Known Continuous Transfer Function Model (General SISO Array).vi .............. 47
SI Estimate Partially Known State-Space Model (MIMO Array).vi ...................................................... 49
Recursive Model Estimation VIs ............................................................................................................ 50
SI Model Simulation VIs ........................................................................................................................ 58
SI Model Prediction VIs ......................................................................................................................... 65
si_Residual Analysis.vi .......................................................................................................................... 70
si_Convert to Transfer Function Model.vi ............................................................................................. 72
si_Convert to State Space Model.vi........................................................................................................ 74
si_Convert to Pole-Zero Model.vi .......................................................................................................... 76
si_Std Freq.vi.......................................................................................................................................... 77
si_Std Root to Polynomial.vi .................................................................................................................. 80
si_Polynomial Root Derivative.vi .......................................................................................................... 82
si_Variance of Real and Imaginary of Roots.vi ...................................................................................... 84
si_Confidence Region of Polynomial.vi ................................................................................................. 85
SI Average Theta and Covariance.vi ...................................................................................................... 87
SI Generate Pseudo-Random Binary Sequence.vi .................................................................................. 88
Gauss-Newton Minimization Method .................................................................................................... 89
Instrumental Variable Method ................................................................................................................ 92
SI Estimate Discrete Transfer Function Model from FRF (SISO arbitrary).vi ....................................... 93
SI Estimate Continuous Transfer Function Model from FRF (SISO arbitrary).vi ................................. 96
SI Estimate State-Space Model from FRF.vi ......................................................................................... 97
National Instruments Corporation
System Identification Toolkit
System Identification Palettes, VIs, and SubVIs

The following table lists the System Identification VIs and the corresponding subVIs that implement
the algorithms.
Palette
Data
Preprocessing
Nonparametric
Model
Estimation
FrequencyDomain Model
Estimation
Parametric
Model
Estimation
VI
SI Remove Trend
SI Detect Feedback
SI Estimate Impulse Response
SubVI
si_Piecewise Linear Fitting.vi
SI Detect Feedback (Array).vi
si_Correlation Analysis.vi
SI Estimate Impulse Response (Least Squares
Array).vi
si_Frequency Analysis.vi
SI Estimate Frequency
Response
SI Estimate Transfer Function
Model from FRF
SI Estimate Discrete Transfer Function Model

from FRF (SISO arbitrary).vi
SI Estimate Continuous Transfer Function
Model from FRF (SISO arbitrary).vi
SI Estimate State-Space Model from FRF.vi
SI Estimate State-Space Model

from FRF
SI Estimate AR Model
si_AR (FB, LS, YW and PC).vi

si_AR Burg Method.vi
si_Est ARX Model (SISO).vi
si_Est ARX Model (MISO).vi
si_Est ARX Model (MIMO).vi
si_Est ARMAX Model (SISO).vi
si_Est ARMAX Model (MISO).vi
si_Est OE Model (SISO).vi
si_Est OE Model (MISO).vi
si_Est BJ Model (SISO).vi
si_Est BJ Model (MISO).vi
si_Est GL Model (SISO).vi
si_Est GL Model (MISO).vi
SIMPCA
si_Impulse Realization (MIMO).vi
SI Estimate Continuous Transfer Function
Model (SISO Array).vi
SI Estimate Discrete Transfer Function Model
(SISO Array).vi
SI Estimate User-Defined Model (MIMO
Array).vi
SI Estimate Partially Known Continuous
Transfer Function Model (General SISO
Array).vi
SI Estimate Partially Known State-Space Model
(MIMO Array).vi
Recursive model estimation
SI Estimate ARX Model

SI Estimate ARMAX Model
SI Estimate OE Model
SI Estimate BJ Model
SI Estimate GL Model
SI Estimate State-Space Model
SI Estimate Transfer Function
Model
Partially Known
Model
Estimation
Recursive
Model
Estimation
Model
SI Estimate User-Defined
Model
SI Estimate Partially Known
Continuous Transfer Function
Model
SI Estimate Partially Known
State-Space Model
SI Recursively Estimate AR
Model
SI Recursively Estimate ARX
Model
SI Recursively Estimate
ARMAX Model
SI Recursively Estimate OE
Model
SI Recursively Estimate BJ
Mode
SI Recursively Estimate StateSpace Model
SI Model Conversion
si_Convert to Transfer Function Model.vi
ni.com
Conversion
SI Convert to Models of CDT
Model Analysis
SI Bode Plot
SI Nyquist Plot
SI Pole-Zero Plot
Model
Validation
si_Convert to State Space.Model.vi

si_Convert to Pole-Zero Model.vi
si_Std Freq.vi
si_Std Root to Polynomial.vi
si_Polynomial Root Derivative.vi
si_Variance of Real and Imaginary of Roots.vi
si_Confidence Region of Polynomial.vi
si_GL Model Sim (MISO).vi
si_Std Response To Model.vi
si_ARX Model Sim (MIMO).vi
si_State-Space (MIMO) core.vi
si_Transfer Function Sim Core.vi
si_GL Model Predict (MISO).vi
si_SS Model Predict vi
si_SS State Response.vi
si_Residual Analysis.vi
SI Average Theta and Covariance.vi
SI Generate Pseudo-Random Binary
Sequence.vi
SI Model Simulation
SI Model Prediction
Utilities
SI Model Residual Analysis

SI Merge Models
SI Generate Pseudo-Random
Binary Sequence
si_Piecewise Linear Fitting.vi

1.
Introduction
This section includes the algorithms and references used by the si_Piecewise Linear Fitting.vi. The
si_Piecewise Linear Fitting.vi is the core VI used to remove a piecewise linear trend from the
input-output data of a system.
2.
Function Description
The purpose of this function is to find the piecewise linear trend of a signal. A signal with a
piecewise linear trend is represented by
s0 sN 1 .
The piecewise points include 0, x1, , xL,
N-1.
3.
APIs
Para Name
signal with
In/Out
Type
Description
In
1D DBL Array
The signal to be fitted
In
1D I32 Array
The piecewise points of the linear fitting
Out
1D DBL Array
Fitted signal
Out
1D DBL Array
The best fit piecewise linear signal
trend
piecewise
points
Signal
(without
trend)
best fit
4.
slopes
Out
1D DBL Array
The piecewise slope
intercepts
Out
1D DBL Array
The piecewise intercept
error
Out
I32
Algorithms
If
y0, y1 yL , yN 1 are the best fit signal values at the piecewise points0, x1, , xL, N-1,
respectively, the best fit value of
sk
between
xi
and
xi 1
is:
x k
k xi
sk yi i 1
yi 1
, xi k xi 1
xi 1 xi
xi 1 xi
(1)
Therefore, you can obtain the piecewise linear trend, or the best fit values, of the signal if you
know y0, y1 yL , yN 1 . Using the least square criterion, you can obtain
y0, y1 yL , yN 1
by
solving the following linear equations:
AY S
where
Y [ y0, y1 yL , y N 1 ]T , S [s0, s1 , s N 1 ]T , and
a00

a0m0
0
0
A
0
0
0
0
0
where
xi
(2)
aik
and
xi 1
xi1 k
xi 1 xi
that satisfies
bik
b00
b0m0
0
0
0
0
0
0
a10
a1m1
b10
b1m1
0
0
0
0
0
0
0
0
0
0
k xi
xi1 xi
0
0
0
bL1,0

bL1,mL1
0
0
0
aL1,0
aL1,mL1
(3)
and mi represents the number of signals between
m0 ... mL1 N . x0 0
and
xN 1 N 1 .
By solving Equation 2, you can calculate the best fit value at the piecewise points,
y0, y1 yL , yN 1 . Then, by multiplying the matrix A by the vector Y, you can calculate the best
fit signal.
ni.com
SI Detect Feedback (Array).vi

1.
Introduction
This section includes the algorithms and references used by the SI Detect Feedback (Array).vi. The
SI Detect Feedback (Array).vi is the core VI used to detect feedback in a system. The SI Detect
Feedback (Array).vi is located in the labview\vi.lib\addons\System Identification\Data
Preprocess.llb directory.
2.
The purpose of this VI is to determine whether data is collected from a closed-loop system using
estimated impulse responses. The SI Detect Feedback (Array).vi uses the least squares method to
estimate impulse responses. Refer to the SI Estimate Impulse Response (Least Squares Array).vi
for more information about estimating impulse responses.
3.
APIs
Para Name
In/Out
Type
Description
stimulus signal
In
1D DBL Array
The input signal of a system
response signal
In
1D DBL Array
The output signal of a system
impulse response
In
Cluster
The number of points for which t<0 and
length
t>=0, represented as
N t 0
and N t 0 ,
respectively.
confidence level
In
Numeric
Determines the upper and lower limits of
(%)
the impulse response
feedback exists?
Out
Boolean
impulse response
Out
XY graph
graph
4.
Algorithms
1)
Estimate the impulse responses, h(k), as well as the standard deviation, std(k), of the impulse
responses.
2)
In an open-loop system, the current input signal, u(k), and the previous output signal, y(k+j),
j<0, do not correlate with each other. Thus, the impulse response h(j) at negative lags (j<0) is
zero. In a closed-loop system, the previous output signal correlates with the current input
signal because of the feedback path in the system. Hence, some values of the impulse
response at negative lags are nonzero. In other words, when the impulse responses of a
system are nonzero at negative lags, feedback exists in the system.
However, signals might contain noise, so the estimated impulse responses at negative lags
might not be zero even in an open-loop system. When h(j) lies in the range from - N*std(-j)
to N*std(-j), where N is determined by the confidence level, h(j) is considered to be zero. If

h(j) does not lie in this range, h(j) is considered to be nonzero, or significant.
si_AR (FB, LS, YW and PC).vi

1.
Introduction
This section includes the algorithms and references used by the si_AR (FB, LS, YW and PC).vi. The
si_AR (FB, LS, YW and PC).vi is the core VI used to estimate AR models. The si_AR (FB, LS,
YW and PC).vi implements the forward-backward, least-square, Yule-Walker, and principle
components methods for AR model estimation.
2.
The Auto-Regression (AR) model is defined as:
A(q) y(n) e(n)

where
Na
A(q) 1 ai q i , e(n) is
(1)
white noise, and y(n) is the signal. q is the backward shift
i 1
operator, which means:
q i y(n) y(n i)
The purpose of this VI is to estimate A(q) given the signal y(t) so that Equation 1 holds.
3.
APIs
Para Name
In/Out
Type
Description
signal
In
1D DBL Array
The signal to be fitted with Equation 1
AR order
In
I32
Number of auto-regression coefficients, Na
method
In
U16 Text Ring
Select the method to estimate the AR model.

The methods include:
(1) Forward-Backward (FB)
(2) Least-Square (LS)
(3) Yule-Walker (YW)
(4) Principal Component (PC)
AR
Out
1D DBL Array
[1, a1, a2 ,....,aNa ]
std
Out
1D DBL Array
Standard deviations of
covariance
Out
2D DBL Array
Covariance matrix of
noise
Out
1D DBL Array
e(n)
noise
Out
DBL
Variance of e(n)
coefficients
[1, a1, a2 ,....,aNa ]
[a1 , a2 ,....,aNa ]
variance
ni.com
error
4.
Out
I32
Error code
Algorithms
Forward-Backward Method
The AR coefficients can be estimated by solving the following linear equations:
M f
m f
a
M b
mb
(2)
where
y(0)
y(na 1) y(na 2)
y (n )
y(na 1)
y(1)
a
Mf

y( N 2) y( N 3) y( N na 1)
y(2)
y(na )
y(1)
y(2)
y(3)
y(na 1)
Mb
y( N na ) y( N na 1) y( N 1)
y(0)
a1
y(na )
a
y(n 1)
y(1)
2
a
mf
, mb
, a

y( N 1)
y( N na 1)
ana
Least Squares Method
y(0)
y(na 1) y(na 2)
a1
y(na )
a
y (n )
y(na 1)
y(1)
a
2 y(na 1)

y( N 1)
y( N 2) y( N 3) y( N na 1) na
(1)
where N is the signal length.

Yule-Walker Method
M f
m f
a
M b
mb
(3)
where
0
y(0)
y(1)
y(0)
0
Mf
y( N na ) y( N na 1) y( N 1)
y(2) y(na )
y(1)
y(2)
y(3) y(na 1)
Mb

0
y( N 1) 0
a1
y(1)
y(0)
a
y(2)
y(1)
2
mf
, mb
, a

ana
y( N 1)
y( N 2)
5.
References
[1] Signal Processing Toolkit User Manual, National Instruments
[2] Guangshu Hu, Digital Signal Processing, Tsing Hua University Press.
si_AR Burg Method.vi

1.
Introduction
This section includes the algorithms and references used by the si_AR Burg Method.vi. The si_AR
Burg Method.vi is the core VI used to estimate an AR model using the Burg method. The si_AR
Burg Method.vi is located in the labview\vi.lib\addons\System Identification\Parametric
Estimation Subs.llb directory.
2.
The Auto-Regression (AR) model is defined as:
A(q) y(n) e(n)

where
(1)
Na
A(q) 1 ai q i , e(n) is white noise, and y(n) is the signal. q is the backward shift
i 1
operator, which means:
q i y(n) y(n i)
The purpose of this VI is to estimate A(q) given the signal y(t) so that Equation 1 holds.
3.
APIs
ni.com
Para Name
In/Out
Type
Description
stimulus
In
2D DBL Array
The input signals of a system
In
1D DBL Array
In
Cluster containing A
A order, B order, and delay are
signal
response
signal
model
parameters
order, B order, and
Na
[N1b, N2b,...., NNb] , and [k1, k2,...., kN ] ,
delay. A order is
respectively.
I32. B order and

delay are 1D I32.
ARX model
Out
Cluster
-
std
Out
Cluster
The standard deviations of coefficients A and B
Out
DBL
Variance of e(n)
Out
2D DBL Array
Out
1D DBL Array
deviations
noise
variance
covariance
matrix
noise
4.
e(n)
Algorithms
Burg Method [Ref 1, page 363]
According to the definition of AR model, you can define:
Na
x f (n) akf (k ) x(n k )

k 1
e f (n) = x(n) - x f (n)

2
r f = E{ e f (n) }
where
x f (n) is the forward prediction vector, e f (n) is the error vector, and a f k are the
coefficients of an AR model.
Correspondingly, you can define:
Na
x b (n na ) akb (k ) x(n na k )
k 1
eb (n) = x(n - na ) - xb (n - na )
2
r b = E{ eb (n) }
The coefficients
a(k ) of an AR model can be computed by following the Levinson recursive
algorithm, where
am (k ) = am- 1 (k ) + kmam- 1 (m - k ), k = 1,..., m - 1
(2)
am (m) = km
(3)
rm = (1- km2 )rm- 1
(4)
N 1
km
2 emf 1 (i)emb 1 (i 1)
, m 1,2,....,na
i m
N 1
| e
i m
f
m1
N 1
(i) | | e
2
i m
b
m1
(i 1) |
(5)
em f (n) = em- 1 f (n) + kmem- 1b (n - 1)

emb (n) = em- 1b (n) + kmem- 1 f (n)
f
(6)
b
0
e0 (n) = e (n) = x(n), m = 1,..., na

5.
References
[1] Guangshu Hu, Digital Signal Processing, Tsing Hua University Press.
si_Est ARX Model (SISO).vi

1.
Introduction
This section includes the algorithms and references used by the si_Est ARX Model (SISO).vi,
which is the core VI to estimate the coefficients of an ARX (SISO) model using the input-output
data of a system. The si_Est ARX Model (SISO).vi is located in the labview\vi.lib\addons\System
Identification\Parametric Estimation Subs.llb directory.
2.
The ARX (SISO) model is described as
A(q) y(n) B(q)u(n k ) e(n)
where A(q) 1
Na
Nb 1
i 1
i 0
(1)
ai qi , B(q) bi qi , k is the delay of the system, and u(n), y(n), and
e(n) are the input, output, and disturbance of a system, respectively. q is the backward shift
operator, which means
q i y(n) y(n i)
The purpose of this VI is to estimate the coefficients
[a1 , a2 ,....,aNa ] and [b0 , b1 ,....,bNb 1 ]
using the input-output data of a system.
10
ni.com
3.
APIs
Para Name
In/Out
Type
Description
stimulus
In
1D DBL Array
In
1D DBL Array
In
A order, B order, and delay are Na ,
signal
response
signal
model
parameters
order, B order, and
Nb ,
and k,
respectively
delay. A order, B
order, and delay are
I32.
ARX model
Out
Cluster
std
Out
Cluster
The standard deviations of A coefficients and B
deviations
noise
coefficients
Out
DBL
Variance of e(n)
Out
2D DBL Array
Out
1D DBL Array
variance
covariance
matrix
noises
4.
e(n)
Algorithms
The least square solution for the ARX model estimation is [Equation 7.34, Ref 1]:
NLS [
N
1 N
T
1 1
(
t
)
(
t
)
]
(t) y(t)
N t 1
N t 1
(2)
where
(t) [ y(t 1) y(t 2).... y(t Na ) u(t) u(t 1)...u(t Nb 1)]T

You can rewrite Equation 2 as the solution of the linear equations:
AX Y
(3)
where
a1
T ( p)
y ( p)
y( p 1)
aNa
( p 1)
A
, X b0 , and Y
y( N )
( N )
bNb 1
5.
References
11
[1] Ljung, L. 1999. System Identification Theory for the User. 2nd ed. Prentice Hall.
si_Est ARX Model (MISO).vi

1.
Introduction
This section includes the algorithms and references used by the si_Est ARX Model (MISO).vi,
which is the core VI to estimate the coefficients of an ARX (MISO) model based on the inputoutput data of a system. The si_Est ARX Model (MISO).vi is located in the
labview\vi.lib\addons\System Identification\Parametric Estimation Subs.llb directory.
2.
An ARX (MISO) model is described as

where A(q) 1
(1)
Na
ai qi , B(q) [B1 (q), B2 (q),....,BN (q)] and Bi (q)

i 1
Nib 1
b q
j 0
ij
u(n k) [u1 (n k1 ),u2 (n k2 ),....,uN (n kN )]T , N is the number of input channels,

and q is the backward shift operator, which means
q i y(n) y(n i) .
k is the delay of the system.
3.
APIs
Para Name
In/Out
Type
Description
stimulus
In
2D DBL Array
In
1D DBL Array
In
signal
response
signal
model
parameters
order, B order, and
Na
[N1b, N2b,...., NNb] , and [k1, k2,...., kN ] ,
delay. A order is
respectively.
I32. B order and

delay are 1D I32.
ARX model
Out
Cluster
std
Out
Cluster
Out
DBL
Variance of e(n)
Out
2D DBL Array
deviations
noise
variance
covariance
12
ni.com
matrix
noise
4.
Out
1D DBL Array
e(n)
Algorithms
Similar to the algorithm of the si_Est ARX Model (SISO).vi.
5.
References
si_Est ARX Model (MIMO).vi

1.
Introduction
This section includes the algorithms and references used by the si_Est ARX Model (MIMO).vi,
which is the core VI to estimate the coefficients of the an ARX (MIMO) model based on the inputoutput data of a system. The si_Est ARX Model (MIMO).vi is located in the
2.
The ARX (MIMO) model is described as

where
Na
A(q) I Ai q
i 1
Nb 1
B(q) Bi q i
i 0
u(n k) [u1 (n k1 ),u2 (n k2 ),....,uN (n kN )]T
(1)
y(n) [ y1 (n), y2 (n),....,yM (n)]T

, I is the unity matrix.
Ai
and
Bi
are
matrices, N is the number of input channels, M is the number of output channels, k is the delay of
the system, and q is the backward shift operator, which means
q i y(n) y(n i) .
3.
APIs
Para Name
In/Out
Type
Description
stimulus
In
2D DBL Array
In
2D DBL Array
The output signals of a system
In
signal
response
signal
model
parameters
order, B order, and
[ N1b , N 2 b ,....,N Nb ] ,
delay. A order is
and [k1, k2,...., kN ] ,
respectively.
I32. B order and
[ N 1a , N 2 a ,....,N Na ] ,
13
delay are ID I32.

ARX model
Out
Cluster
std
Out
Cluster
Out
2D DBL Array
Variance of e(n)
Out
3D DBL Array
Out
2D DBL Array
deviations
noise
variance
covariance
matrix
noise
4.
e(n)
Algorithms
Similar to the algorithm of the si_Est ARX Model (SISO).vi.
5.
References
si_Est ARMAX Model (SISO).vi

1.
Introduction
This section includes the algorithms and references used by the si_Est ARMAX Model (SISO).vi,
which is the core VI to estimate the coefficients of an ARMAX (SISO) model based on the inputoutput data of a system. The si_Est ARMAX Model (SISO).vi is located in the
2.
The ARMAX (SISO) model is described as
A(q) y(n) B(q)u(n k ) c(q)e(n)

where
(1)
Na
Nb 1
Nc
i 1
i 0
i 1
A(q) 1 ai q i , B(q) bi q i , C(q) 1 ci q i
, u(n), y(n) and e(n) are
the input, output, and disturbance of a system, respectively. q is the backward shift operator, which
means
q i y(n) y(n i) .
k is the delay of the system. The purpose of this VI is to estimate the coefficients [a1, a2,....,aN ] ,
a
[b0, b1,....,bN 1] and [c1, c2,....,cN ] based on the input-output data of a system.
b
c
3.
APIs
Para Name
In/Out
Type
Description
14
ni.com
stimulus
In
1D DBL Array
In
1D DBL Array
In
A order, B order, C order and delay are Na ,
signal
response
signal
model
parameters
order, B order, C
Nb , Nc , and k, respectively.
order and delay. A

order, B order, C
order and delay are
I32.
ARMAX
Out
Cluster
Out
Cluster
model
std
The standard deviations of coefficients A, B,
deviations
noise
and C
Out
DBL
Variance of e(n)
Out
2D DBL Array
Out
1D DBL Array
variance
covariance
matrix
noise
4.
e(n)
Algorithms
The multi-stage method is applied to have a coarse estimation for A(q), B(q), and C(q), and
then the Gauss-Newton minimization method is applied to refine the results of A(q), B(q), and
C(q).
Multi-Stage coarse estimation

This method is generally described on page 337 of Ref 1. Here is the deduction specific for an
ARMAX model.
Let
v(t) = c(q)e(t)
(2)
Then Equation 1 becomes

(3)
A(q)y(t)=B(q)u(t-k)+ v(t)
Since v(t) here is not white Gaussian noise, the System Identification Toolkit applies the
instrumental variable (IV) method to estimate A(q) and B(q). You then can have
v(t) A (q) y(t) B (q)u(t k) C(q)e(t)

Here, as elsewhere,
denotes the estimation of
1
v(t ) e(t )
C(q)
15
(4)
. Thus, Equation 4 can be rewritten as

(5)
Equation 4 can be treated as a high order AR model, though theoretically, it represents an AR

model with infinite orders. However, in practice, this kind of AR model is selected as the
dimension of the system Na + Nb + Nc . With the AR model estimation, we have an estimation
for e(t). Since the
v(t ) and e(t ) are known now, Equation 4 can be rewritten as
v(t) (C(q) 1)e(t) e(t)
(6)
Equation 5 represents one form of ARX model. C(q) can be estimated by using ARX model
v(t ) and e(t ) as the output and input, respectively.
estimation methods with
Standard Deviations of A(q), B(q) and C(q)

Refer to the equations (9.18) and (9.19) on Page 284 of Ref 1.
5.
References
si_Est ARMAX Model (MISO).vi

1.
Introduction
This section includes the algorithms and references used by the si_Est ARMAX Model (MISO).vi,
which is the core VI to estimate the coefficients of an ARMAX (MISO) model based on the inputoutput data of a system. The si_Est ARMAX Model (MISO).vi is located in the
2.
The ARMAX (MISO) model is described as
A(q) y(n) B(q)u(n k ) c(q)e(n)

where A(q) 1
Na
a q
i 1
(1)
B(q) [B1 (q), B2 (q),....,BN (q)] and Bi (q)
Nib 1
b q
j 0
ij
Nc
C(q) 1 ci q i , u(n k) [u1 (n k1 ),u2 (n k2 ),....,uN (n kN )]T , N is the

i 1
number of input channels, and q is the backward shift operator, which means
q i y(n) y(n i) .
ki
is the delay of the i-th input of the system. The purpose of this VI is to estimate the coefficients
[a1 , a2 ,....,aNa ] , [bij ]

3.
and
[c1 , c2 ,....,cNc ] based on the input-output data of a system.
APIs
Para Name
In/Out
Type
Description
16
ni.com
stimulus signal
In
2D DBL Array
response signal
In
1D DBL Array
model
In
A order, B order, C order and delay are
parameters
order, B order, C order

and delay. A order, C
order and delay are I32.
Na
[ N1b , N 2 b ,....,N Nb ] , N c ,
and k,
respectively.
B order is 1D I32.
ARMAX model
Out
Cluster
std deviations
Out
Cluster
The standard deviations of coefficients A,

B, and C
noise variance
Out
DBL
Variance of e(n)
covariance
Out
2D DBL Array
Out
1D DBL Array
matrix
noise
4.
e(n)
Algorithms
Similar to the algorithm of the si_Est ARMAX Model (SISO).vi.
5.
References
si_Est OE Model (SISO).vi

1.
Introduction
This section includes the algorithms and references used by the si_Est OE Model (SISO).vi, which
is the core VI to estimate the coefficients of an output-error (OE) (SISO) model based on the
input-output data of a system. The si_Est OE Model (SISO).vi is located in the
2.
An OE (SISO) model is described as
y(n)
where,
Nb 1
B(q) bi q
i 0
B(q)
u(n k ) e(n)
F (q)
, F (q) 1
Nf
fq
i 1
(1)
, u(n), y(n) and e(n) are the input, output, and
disturbance of a system respectively. q is the backward shift operator, which means
q i y(n) y(n i) .
17
k is the delay of the system. The purpose of this VI is to estimate the coefficients
[b0 , b1 ,....,bNb 1 ] and [ f1 , f 2 ,...., f N f ] based on the input-output data of a system.
3.
APIs
Para Name
In/Out
Type
Description
stimulus
In
1D DBL Array
In
1D DBL Array
In
Cluster containing B order,
B order, F order and delay are
signal
response
signal
model
parameters
F order and delay. They are

I32.
OE model
Out
Cluster
Nb , N f
and k respectively.
B is [0,...0, b0 , b1 ,....,bNb 1 ] , The
number of zeros at the beginning
equals k. F is
std
Out
Cluster
The standard deviations of
deviations
noise
[1, f1 , f 2 ,...., f N f ]
coefficients B and F
Out
DBL
Variance of e(n)
Out
2D DBL Array
Out
1D DBL Array
variance
covariance
matrix
noise
4.
e(n)
Algorithms
The multi-stage method is applied to have a coarse estimation for B(q) and F(q), and then the
Gauss-Newton minimization method is applied to refine the results of B(q) and F(q).
Multi-Stage Coarse Estimation

This method is generally described on page 337 of Ref 1. Here is the deduction specific for OE
model.
Let
v(t) = F(q)e(t)
(2)

F(q)y(t)=B(q)u(t-k)+ v(t)
(3)
Equation 3 is in the form of ARX model. Since v(t) here is not white Gaussian noise, the System
Identification Toolkit applies the instrumental variable (IV) method to estimate F(q) and B(q).
Standard Deviations of B(q) and F(q)
18
ni.com

5.
References
si_Est OE Model (MISO).vi

1.
Introduction
This section includes the algorithms and references used by the si_Est OE Model (MISO).vi,
which is the core VI to estimate the coefficients of an output-error (OE) (MISO) model based on
the input-output data of a system. The si_Est OE Model (MISO).vi is located in the
2.
An OE (MISO) model is described as
y(n)
B(q)
u(n k ) e(n)
F (q)
(1)
where

F (q) [F1 (q), F2 (q),....,FN (q)]
Nib 1
b q
j 0
ij
Nif
and
Fi (q) 1 f ij q j
j 1
B (q)
B(q) B1 (q) B2 (q)
[
,
,..., N ] ,
F (q) F1 (q) F2 (q)
FN (q)
u(n k) [u1 (n k1 ),u2 (n k2 ),....,uN (n kN )]T , N is the number of input channels and q
is the backward shift operator, which means:
q i y(n) y(n i) .
ki
[bij ] and [ fij ]

3.
based on the input-output data of a system.
APIs
Para Name
stimulus
In/Out
Type
Description
In
2D DBL Array
In
1D DBL Array
signal
response
19
signal
model
In
parameters
Cluster containing B order,
B order, F order and delay
F order and delay. They
are [ N1b , N 2b ,....,N Nb ] ,
are 1D I32.
OE model
Out
[ N1 f , N 2 f ,....,N Nf ] ,
and k, respectively.
Cluster
B is a matrix as
0,...0, b10 , b11,....,b1( N1b 1)
0,...0, b20 , b21,....,b2( N2b 1) . The

...
0,...0, bN 0 , bN1 ,....,bN ( N Nb 1)

number of zeros at the beginning of ith
row equals
ki
F is a matrix as
1, f11, f12 ,...., f1N1 f
1, f 21, f 22 ,...., f 2 N2 f
...
1, f N1 , f N 2 ,...., f NN
Nf
std deviations
Out
Cluster

coefficients B and F
noise variance
Out
DBL
Variance of e(n)
covariance
Out
2D DBL Array
Out
1D DBL Array
matrix
noise
4.
e(n)
Algorithms
The multi-stage method is applied to have a coarse estimation for B(q) and F(q), and then the GaussNewton minimization method is applied to refine the results of B(q) and F(q).
This method is generally described on page 337 of Ref 1. Here is the deduction specific for the OE
model.
By multiplying
F max(q)
on both side of Equation 1, this equation is approximated as
F max(q) y(n) B(q)u(n k ) (n)
(2)
where
(n) F max(q)e(n)
Here
F max(q)
is the polynomial with maximum order in all
(3)
Fi (q) . Then Equation 2 can be
considered as an ARX (MISO) model. With the ARX (MISO) model estimation, you can have an
20
ni.com
estimation for
F max(q)
and B(q). Using
F max(q)
and B(q) as filter N(q) and M(q), you can apply
the instrumental variable method to estimate B(q) and F(q) of Equation 1.

Standard Deviations of B(q) and F(q)
Refer to the Equations (9.18) and (9.19) on Page 284 of Ref 1.
5.
References
si_Est BJ Model (SISO).vi

1.
Introduction
This section includes the algorithms and references used by the si_Est BJ Model (SISO).vi, which
is the core VI to estimate the coefficients of a Box-Jenkins (BJ) (SISO) model based on the inputoutput data of a system. The si_Est BJ Model (SISO).vi is located in the
2.
A BJ (SISO) model is described as
y(n)
where
B(q)
C(q)
u(n k )
e(n)
F (q)
D(q)
(1)
Nb 1
Nf
Nc
Nd
i 0
i 1
i 1
i 1
B(q) bi q i , F (q) 1 f i q i , C(q) 1 ci q i , D(q) 1 di q i
and q is the backward shift operator, which means:
q i y(n) y(n i) .
k is the delay of the system. The purpose of this function is to estimate the coefficients
[b0 , b1 ,....,bNb 1 ] , [ f1 , f 2 ,...., f N f ] , [c1 , c2 ,....,cNc ]
and
[d1, d2 ,....,dNd ]
based on the
input-output data of a system.

3.
APIs
Para Name
stimulus
In/Out
Type
Description
In
1D DBL Array
In
1D DBL Array
In
Cluster
B order, F order, C order, D order and
signal
response
signal
model
parameters
delay are
21
Nb
Nf
Nc , Nd
and k,
respectively.
BJ model
Out
Cluster
B is [0,...0, b0 , b1 ,....,bN
b 1
] , The

equals k. F is
is
[1, f1 , f 2 ,...., f N f ] .
[1, c1, c2 ,....,cNc ] and D is
[1, d1, d2 ,....,d Nd ]

std deviations
Out
Cluster

coefficients B, F, C, and D
noise variance
Out
DBL
Variance of e(n)
covariance
Out
2D DBL Array
Out
1D DBL Array
matrix
noise
4.
e(n)
Algorithms
The multi-stage method is applied to have a coarse estimation for B(q), F(q), C(q) and D(q), and
then the Gauss-Newton minimization method is applied to refine the results of B(q), F(q), C(q)
and D(q).
This method is generally described on page 337 of Ref 1. Here is the deduction specific for BJ
model.
Let
(n) F (q)
C(q)
e(n)
D(q)
(2)
F (q) y(n) B(q)u(n k ) (n)
(3)
You then apply the instrumental variable (IV) method to estimate B(q) and F(q) and have the
following equation:
v(n) y(n)
Here, as elsewhere,
B (q)
C(q)
u(n k )
e(n)
D(q)
F (q)
. If
Nc 0 ,
(4)
Equation 4 can be re-written as
D(q)v(n) e(n)
(5)
Equation 5 can be treated as an AR model. With the AR model estimation, you have an
estimation for D(q). If
Nc 0 ,
Equation 4 is rewritten as
D(q)
v(n) e(n)
C(q)
22
(6)
ni.com
Then a high order AR model is applied to estimate
e(n) .
Since the
v(n) and e(n) are known
now, Equation 6 can be rewritten as
D(q)v(n) (C(q) 1)e(n) e(n)
(7)
Equation 7 is one form of the ARX model. C(q) and D(q) can be estimated by using ARX model
v(n)
e(n) as the output and input, respectively.
and
Standard Deviations of B(q), F(q), C(q) and D(q)

5.
References
si_Est BJ Model (MISO).vi

1.
Introduction
This document includes the algorithms and references used by the si_Est BJ Model (MISO).vi,
which is the core VI to estimate the coefficients of a Box-Jenkins (BJ) (MISO) model based on the
input-output data of a system. The si_Est BJ Model (MISO).vi is located in the
2.
A BJ (MISO) model is described as:
B(q)
C(q)
u(n k )
e(n)
F (q)
D(q)
y(n)
(1)
where

F (q) [F1 (q), F2 (q),....,FN (q)]
Nib 1
b q
j 0
ij
Nif
and
Fi (q) 1 f ij q j
j 1
B (q)
B(q) B1 (q) B2 (q)
[
,
,..., N ] ,
F (q) F1 (q) F2 (q)
FN (q)

Nc
C(q) 1 ci q
i 1
Nd
D(q) 1 di q i
i 1
23
q i y(n) y(n i) .
ki
[bij ] , [ fij ] , [c1 , c2 ,....,cNc ] , and [d1, d2 ,....,d Nd ] based on the input-output data of a system.
3.
APIs
Para Name
stimulus
In/Out
Type
Description
In
2D DBL Array
In
1D DBL Array
In
Cluster
B order, F order, C order, D order and
signal
response
signal
model
parameters
delay are
[ N1b , N 2 b ,....,N Nb ] ,
[ N1 f , N 2 f ,....,N Nf ] , N c , N d
[ k1 , k 2 ,....,k N ] ,
BJ model
Out
Cluster
and
respectively.
B is a matrix as
0,...0, b10 , b11,....,b1( N1b 1)
0,...0, b20 , b21,....,b2( N2b 1) . The

...
0,...0, bN 0 , bN1 ,....,bN ( N Nb 1)

row equals
ki
F is a matrix as
1, f11, f12 ,...., f1N1 f
1, f 21, f 22 ,...., f 2 N2 f
...
1, f N1 , f N 2 ,...., f NN
Nf
C is
[1, c1, c2 ,....,cNc ] and D is
[1, d1, d2 ,....,d Nd ]

std deviations
Out
Cluster

coefficients B, F, C, and D
noise variance
Out
DBL
covariance
Out
2D DBL Array
Out
1D DBL Array
Variance of e(n)
-
matrix
noise
e(n)
24
ni.com
4.
Algorithms
The multi-stage method is applied to have a coarse estimation for B(q), F(q), C(q) and D(q), and
then the Gauss-Newton minimization method is applied to refine the results of B(q), F(q), C(q) and
D(q).
This method is generally described on page 337 of Ref 1. Here is the deduction specific for BJ
model.
By multiplying
F max(q)
on both side of Equation 1, Equation 1 is approximated as
F max(q) y(n) B(q)u(n k ) (n)
(2)
where
(n) F max(q)
Here
F max(q)
C(q)
e(n)
D(q)
Fi (q) . Then Equation 2 can be
considered as an ARX (MISO) model. With the ARX (MISO) model estimation, you can have an
estimation for
F max(q)
and B(q). Using
F max(q)
and B(q) as filter N(q) and M(q), then you
can apply the Instrumental Variable method to estimate B(q) and F(q) of Equation 1 and have:
v(n) y(n)
Here, as elsewhere,
B (q)
C(q)
u(n k )
e(n)
D(q)
F (q)
. If
Nc 0 ,
(3)
Equation 3 can be re-written as
D(q)v(n) e(n)
(4)
Equation 4 can be treated as an AR model. With the AR model estimation, you have an estimation
for D(q). If
Nc 0 ,
D(q)
v(n) e(n)
C(q)
e(n) .
Since
(5)
v(n) and e(n) are known now,
Equation 5 can be rewritten as
D(q)v(n) (C(q) 1)e(n) e(n)
(6)
Equation 6 is one form of ARX model. Then C(q) and D(q) can be estimated by using ARX model
v(n)
and
Standard Deviations of B(q), F(q), C(q) and D(q)

Refer to the Equation (9.18) and (9.19) on Page 284 of Ref 1.
5. References
25
si_Est GL Model (SISO).vi

1.
Introduction
This section includes the algorithms and references used by the si_Est GL Model (SISO).vi, which
is the core VI to estimate the coefficients of a general-linear (GL) (SISO) model based on the
input-output data of a system. The si_Est GL Model (SISO).vi is located in the
labview\vi.lib\addons\System Identification\Parametric Estimation Subs.llb.
2.
A GL (SISO) model is described as:
A(q) y(n)
where
B(q)
C(q)
u(n k )
e(n)
F (q)
D(q)
(1)
Na
Nb 1
Nf
Nc
i 1
i 0
i 1
i 1
A(q) 1 ai q i , B(q) bi q i , F (q) 1 f i q i , C(q) 1 ci q i

Nd
D(q) 1 di q i
i 1
q i y(n) y(n i) .
k is the delay of the system. The purpose of this VI is to estimate the coefficients
,
[a1 , a2 ,....,aNa ]
[b0 , b1 ,....,bNb 1 ]
[ f1 , f 2 ,...., f N f ]
and
[c1 , c2 ,....,cNc ]
[d1, d2 ,....,dNd ] based on the input-output data of a system.

3.
APIs
Para Name
stimulus
In/Out
Type
Description
In
1D DBL Array
In
1D DBL Array
In
Cluster
A order, B order, F order, C order, D
signal
response
signal
model
parameters
order and delay are

Nd
GL model
Out
Cluster
Na
Nb
Nf
Nc ,
and k, respectively.
A is
[1, a1 , a2 ,....,aNa ] . B
is [0,...0, b0 , b1 ,....,bNb 1 ] , The

equals k. F is
26
[1, f1 , f 2 ,...., f N f ] .
ni.com
is
[1, c1, c2 ,....,cNc ] and D is
[1, d1, d2 ,....,d Nd ]

std deviations
Out
Cluster

coefficients A, B, F, C, and D
noise variance
Out
DBL
Variance of e(n)
covariance
Out
2D DBL Array
Out
1D DBL Array
matrix
noise
4.
e(n)
Algorithms
The multi-stage method is applied to have a coarse estimation for A(q), B(q), F(q), C(q) and D(q),
and then the Gauss-Newton minimization method is applied to refine the results of A(q), B(q), F(q),
C(q) and D(q).
Multi-Stage coarse estimation
This method is generally described on page 337 of Ref 1. Here is the deduction specific for GL
model.
Let
(n) F (q)
C(q)
e(n)
D(q)
(2)
A(q)F (q) y(n) B(q)u(n k ) (n)
(3)
You then can apply the instrumental variable (IV) method to estimate B(q) and A(q)F(q). Let
(n)
C(q)
e(n) , then Equation 1 becomes
D(q)
A(q) y(n)
B(q)
u(n k ) (n)
F (q)
(4)
You can approximately consider Equation 4 as an ARX model, whose B order is high
(
Nb
Nf
Nc
+ N d ). Then by applying the instrumental variable method, you get A(q). Since
A(q)F(q) and A(q) are known now, you can compute F(q) (=A(q)F(q)/A(q)). So far you already
have estimation for A(q), B(q), and F(q). Substituting them into Equation 1, you have
B (q)
C(q)
v(n) A(q) y(n)
u (n k )
e(n)
D(q)
F (q)
Here, as elsewhere,
. If
Nc 0 ,
D(q)v(n) e(n)
27
(5)
(6)
Equation 6 can be treated as an AR model. With the AR model estimation, you can have an
estimation for D(q). If
Nc 0 ,
D(q)
v(n) e(n)
C(q)
e(n) .
Since
(7)
D(q)v(n) (C(q) 1)e(n) e(n)
(8)
Equation 8 is one form of ARX model. C(q) and D(q) can be estimated by using ARX model
v(n)
and
Standard Deviations of A(q), B(q), F(q), C(q) and D(q)

Refer to Equations (9.18) and (9.19) on Page 284 of Ref 1.
5.
References
si_Est GL Model (MISO).vi

1.
Introduction
This section includes the algorithms and references used by the si_Est GL Model (MISO).vi,
which is the core VI to estimate the coefficients of a general-linear (GL) (MISO) model based on
the input-output data of a system. The si_Est GL Model (MISO).vi is located in the
2.
A GL (MISO) model is described as:
A(q) y(n)
where
Na
A(q) 1 ai q i
B(q)
C(q)
u(n k )
e(n)
F (q)
D(q)
(1)
i 1

F (q) [F1 (q), F2 (q),....,FN (q)]
Nib 1
b q
j 0
ij
Nif
and
Fi (q) 1 f ij q j
j 1
B (q)
B(q) B1 (q) B2 (q)
[
,
,..., N ] ,
F (q) F1 (q) F2 (q)
FN (q)
28
ni.com

Nc
Nd
i 1
i 1
C(q) 1 ci q i , D(q) 1 di q i
q i y(n) y(n i) .
ki
[a1 , a2 ,....,aNa ] , [bij ] , [ fij ] , [c1 , c2 ,....,cNc ] ,
and
[d1 , d2 ,....,d Nd ]
based on the input-
output data of a system.

3.
APIs
Para Name
stimulus
In/Out
Type
Description
In
2D DBL Array
In
1D DBL Array
In
Cluster containing A order,
A order, B order, F order, C order, D
B order, F order, C order, D
order and delay are
signal
response
signal
model
parameters
order and delay. B order and

F order are 1D I32. A order ,
C order, D order and delay
Na
[ N1b , N 2 b ,....,N Nb ] , [ N1 f , N 2 f ,....,N Nf ] , N c ,

Nd
and
[ k 1 , k 2 ,....,k N ] ,
respectively.
are I32.
GL model
Out
Cluster
A is
[1, a1 , a2 ,....,aNa ] .
B is a matrix as
0,...0, b10 , b11,....,b1( N1b 1)
0,...0, b20 , b21,....,b2( N2b 1) . The

...
0,...0, bN 0 , bN1 ,....,bN ( N Nb 1)

row equals
ki
F is a matrix as
1, f11, f12 ,...., f1N1 f
1, f 21, f 22 ,...., f 2 N2 f
...
1, f N1 , f N 2 ,...., f NN
Nf
C is
29
[1, c1, c2 ,....,cNc ] and D is

[1, d1, d2 ,....,d Nd ]

std
Out
Cluster
The standard deviations of coefficients
deviations
noise
A, B, F, C, and D
Out
DBL
Variance of e(n)
Out
2D DBL Array
Out
1D DBL Array
variance
covariance
matrix
noise
4.
e(n)
Algorithms
The multi-stage method is applied to have a coarse estimation for A(q), B(q), F(q), C(q) and D(q),
and then the Gauss-Newton minimization method is applied to refine the results of A(q), B(q), F(q),
C(q) and D(q).
This method is generally described on page 337 of Ref 1. Here is the deduction specific for GL
(MISO) model.
Let
(n) F (q)
and set
C(q)
e(n)
D(q)
Fi (q) 1, then Equation 1 becomes

A(q)F max(q) y(n) B(q)u(n k ) (n)
where
(2)
F max(q)
estimation methods to estimate B(q) and
A(q)F max(q) .
(3)
Fi (q) . You can apply ARX model
Since you can approximately consider
Fi (q) 1 in this initial estimation step, Equation 1 can be rewritten as

A(q) y(n) B(q)u(n k ) (n)
Using
(4)
A(q)F max(q) and B(q) as filter and then applying the Instrumental Variable method, you
have the estimation for
A(q)
and
B(q) . Substituting them into Equation 1, you can have
B (q)
C(q)
v(n) A(q) y(n)
u (n k )
e(n)
D(q)
F (q)
Here, as elsewhere,
denote the estimation of
. If
Nc 0 ,
(5)
D(q)v(n) e(n)
(6)
Equation 6 can be treated as an AR model. With the AR model estimation, you have an estimation
for D(q). If
Nc 0 ,
Equation 5 is re-written as
D(q)
v(n) e(n)
C(q)
30
(7)
ni.com
e(n) .
Since the
D(q)v(n) (C(q) 1)e(n) e(n)
(8)
Equation 8 is one form of ARX model. Then C(q) and D(q) can be estimated by using ARX model
v(n)
and
Standard Deviations of A(q), B(q), F(q), C(q) and D(q)

Refer to Equations (9.18) and (9.19) on Page 284 of Ref 1.
5.
References
SI Estimate Continuous Transfer Function Model

(SISO Array).vi
1.
Introduction
This section includes the algorithms and references used by the SI Estimate Continuous Transfer
Function Model (SISO Array).vi. The SI Estimate Continuous Transfer Function Model (SISO
Array).vi is the core VI used to estimate continuous transfer functions. The SI Estimate
Continuous Transfer Function Model (SISO Array).vi is located in the
labview\vi.lib\addons\System Identification\Parametric Estimation.llb directory.
2.
The continuous transfer function model can be represented as
y(k ) H (s)u(k ) e(k )
(1)
where u(k) and y(k) are the input and output of a system, e(k) is the additive noise, and H(s) is
represented by the following equation.
H ( s)
b0 b1s... b j s j ... bm s m sTd

e , mn
a0 a1s... ai s i ... an s n
where m is the numerator order, n is the denominator order, and
Td
(2)
is the time delay. The purpose
of this VI is to estimate the transfer function H(s).

3.
APIs
Para Name
In/Out
Type
Description
stimulus signal
In
1D DBL Array
response signal
In
1D DBL Array
orders of
In
Cluster
m is the numerator order and n is the
31
transfer function
denominator order
model
delay initial
In
DBL
The initial guess of
sampling rate
In
DBL
system model
Out
Object containing
guess (s)
out
information about this
model
coefficients of
Out
Cluster
Numerator is b0, b1,...,bm ,
transfer function
Denominator is a0 , a1,...,an
model
delay (s)
4.
Td
Out
DBL
The estimated
Td
Algorithms
This VI uses three steps to estimate the continuous transfer function H(s).
1)
Estimate an equivalent discrete transfer function.
Using the zero-order-hold method, the continuous transfer function
H(s) can be converted to the
following discrete transfer function.
H ' z z Td / dtGz
(3)
v0 v1 z .... vn z n
Gz
u0 u1 z ... un z n
(4)
where
dt is the sampling interval of the stimulus and response signals.
vn
might be zero when
m n.
Equation 4 uses a slightly higher numerator order to include all the possible cases.
The first step is to estimate the discrete transfer function
Gz . Refer to the SI Estimate Discrete
Transfer Function Model (SISO Array).vi for more information about estimating the discrete
transfer function model.
2)
Convert the estimated discrete transfer function to a continuous transfer function.
Using the zero-order-hold method, you can convert
Gz to a continuous transfer function. The
converted model might have a higher numerator order than m. The SI Estimate Continuous
Transfer Function Model (SISO Array).vi truncates unnecessarily high order coefficients, and the
coefficients you obtain are the initial estimates for minimization in step 3.
3)
Use the Gauss-Newton minimization method to refine the initial estimation.
The initial guess of
Td
and the initial estimates of
ai and bj
are tuned by the Gauss-Newton
optimization method to minimize the following cost function:
32
ni.com
VN
1 N 1 2
e (k )
N k 1 2
(5)
SI Estimate Discrete Transfer Function Model (SISO

Array).vi
1.
Introduction
This section includes the algorithms and references used by the SI Estimate Discrete Transfer
Function Model (SISO Array).vi. The SI Estimate Discrete Transfer Function Model (SISO
Array).vi is the core VI used to estimate discrete transfer functions with the direct identification
approach. The SI Estimate Discrete Transfer Function Model (SISO Array).vi is located in the
labview\vi.lib\addons\System Identification\Parametric Estimation.llb directory.
2.
The discrete transfer function model can be represented as
y(k ) H ( z)u(k ) e(k ) ,
(1)
where u(k) and y(k) are the stimulus and response signals, e(k) is the additive noise, and H(z) is
b0 b1 z... bm z m
H ( z)
, m n
a0 a1 z... an z n
(2)
where m is the numerator order and n is the denominator order. The purpose of this VI is to
estimate the transfer function H(z) using the stimulus and response signals of a system.
3.
APIs
Para Name
In/Out
Type
Description
stimulus signal
In
1D DBL Array
response signal
In
1D DBL Array
orders of
In
Cluster
transfer function
denominator order
model
sampling rate
In
DBL
system model
Out
Object containing
out

model
coefficients of
Out
b0 ,b1,...,bm ,
Denominator is a0 , a1 ,...,an
Cluster
Numerator is
transfer function
model
33
4.
Algorithm
The model shown in Equation 1 is one type of output-error (OE) model. Therefore, you can
estimate the transfer function H(z) by estimating an equivalent OE model. You can obtain the
order of this OE model by expressing H(z) in terms of the negative power of z. The following
equation shows the relationship between the discrete transfer function and an OE model.
H z 1 z ( nm)
bm ... b1 z1m b0 z m
an ... a1 z1n a0 z n
(3)
Using Equation 3, you can find that the B order, F order, and delay of this equivalent OE model
are m+1, n, and n-m, respectively.
You also can apply the indirect identification and joint input-output identification approaches to
estimate a transfer function model. Refer to the LabVIEW System Identification Toolkit User
Manual for information about the indirect identification and joint input-output identification
approaches. With the joint input-output identification approach, any open-loop model estimation
method can be used to estimate the transfer function between the reference signal and the stimulus
signal. The System Identification Toolkit uses a long FIR model to estimate this transfer function.
[Ref 1]
5.
References
[1] Paul M.J.Van Den Hof, Ruud J.P.Schrama: An Indirect Method for Transfer Function
Estimation from Closed Loop Data, Automatica, Vol. 29, No. 6, 1993, pp. 1523-1527.
SI Estimate User-Defined Model (MIMO Array).vi

1.
Introduction
This section includes the algorithms and references used by the SI Estimate User-Defined Model
(MIMO Array).vi. The SI Estimate User-Defined Model (MIMO Array).vi is the core VI used to
estimate user-defined models. The SI Estimate User-Defined Model (MIMO Array).vi is located in
the labview\vi.lib\addons\System Identification\Parametric Estimation.llb directory.
2.
The purpose of this VI is to estimate a user-defined model. You can use a template VI to define a
linear or nonlinear model and then use the SI Estimate User-Defined Model (MIMO Array).vi to
estimate this defined model.
3.
APIs
Para Name
stimulus signal
In/Out
In
Type
Description
2D DBL Array
34
ni.com
response signal
In
2D DBL Array
user defined
In
Path
The file path of the VI that defines the
model
model to be estimated
variables
In
1D Cluster Array
The names and initial guesses of the

variables
data
In
Variant
The static data that the user-defined

function needs at run time
termination
In
Cluster
The stop conditions for model

estimation
response
Out
2D DBL Array
The simulated response signal from
obtained by
the estimated model
estimated model
optimized
Out
1D Cluster Array
variables
number of
The name and optimized value of

each variable
Out
Numeric
The number of times this VI calls the
function calls
user-defined model subVI during the

estimation process
4.
Algorithm
You can estimate a user-defined model by refining the initial estimates using Gauss-Newton
optimization routines. The purpose of optimization is to minimize the following cost function:
1 N 1
1 N 1
VN e 2 (k , ) yk , yk
N k 1 2
N k 1 2
where
and
is the variable vector,
(1)
yk , is the response the user-defined model subVI calculates,
yk is the measured response.
si_Impulse Realization (MIMO).vi

1.
Introduction
This section includes the algorithms and references used by the si_Impulse Realization
(MIMO).vi. The si_Impulse Realization (MIMO).vi is the core VI used to estimate state-space
models from the impulse response of a system. The si_Impulse Realization (MIMO).vi is located
in the labview\vi.lib\addons\System Identification\ Parametric Estimation Subs.llb directory.
2.
State-space models in this VI are defined as:
35
xt 1 Axt But
yt Cxt Dut
(1)
The goal of this VI is to estimate the system matrixes A, B, C, and D using the impulse response
hk , where k>=0.
3.
APIs
Para Name
In/Out
Type
Description
impulse response
In
Cluster
The impulse response of this system
model order
In
I32
The number of states
length
In
I32
The length of the impulse response used to

estimate the model
coefficients of state-
Out
Cluster
The A, B, C, and D matrixes of the state-
space model
4.
space model
Algorithms
The impulse response of this system can be computed as
h0 D , hk CAk 1B ,
k>0
(2)
The Hankel matrix is defined as
hk 1 hk 2 ... hk j 1 hk j
h
...
k 2 hk 3 ...
...
H ij k hk 3
...
hk i ... ... ... hk i j
(3)
Hij k Oi Ak Fj
where
Oi
and
Fj
are the extended observability and controllability matrixes, respectively.
Oi C CA ... CAi1
, F B
T
AB ... A j 1B
From the above equations, you can deduce:
Hij 0 Oi Fj ,
(4)
Hij 1 Oi AFj
(5)
Oi B Hi,1 0 , CFj H1, j 0
(6)
The SVD of
Hij 0 is
H ij 0 U1 U 2 1
V1T
T
T U1 1V1
V2
36
(7)
ni.com
where
contains the n largest singular values, n is the model order, or number of states, and
U1 and V1 are orthogonal matrixes. [Equation 3-7, Ref 1]

From Equations 4 and 7, you can deduce:
1
Oi U1 12 , Fi 12 V1T
[Ref 2]
(8)
From Equations 5, 6, and 8, you can deduce:

1
11 2
A 1 2 U1T Hi, j 1V
B 1 2 U1T Hi,1 0
C H1, j 0V1 1
1
2
The above three equations are the Zeiger-McEwen approximate realization algorithm. [Ref 2]
5.
References
[1] Xmath System Identification Module, Part 1
[2] H.P. Zeiger and A.J.McEwen, Approximate Linear Realization of Given Dimension via Hos
Algorithm, IEEE Transactions on Automatic Control, Volume 19, Issue 2, Apr 1974 Page: 153
Subspace Identification Method via Principal

Component Analysis (SIMPCA)
1.
Introduction
This section includes the algorithms and references used by the si_State Space Model Estimation
(SISO).vi and the si_State Space Model Estimation (MIMO).vi. They are the core VIs to estimate
the coefficients of a state-space model based on the input-output data of a system. The si_State
Space Model Estimation (SISO).vi and the si_State Space Model Estimation (MIMO).vi are
located in the labview\vi.lib\addons\System Identification\Parametric Estimation Subs.llb directory.
2.
The state space model in the System Identification Toolkit is defined as
qxt Axt But Ket

yt Cxt Dut et
The goal of this VI is to find an estimate for the system matrix A, B, C, D, disturbance dynamics
Kalman gain K, and the initial states X0 , based on the stimulus and response data sequence
and
ut
yt .
37
3.
API
For SISO Case
Para Name
In/Out
Type
Description
stimulus
In
1D DBL Array
The stimulus signal into the system
response
In
1D DBL Array
The response output from the system
# of states
In
I32
Number of states
# p,f
In
I32
Estimation horizon
structure
In
Cluster
Information about the D matrix, K matrix and
selector
SS coef
initial state
Out
Cluster
The coefficients of state space model, A, B, C,

and D
Out
1D DBL Array
Kalman gain
X0
Out
1D DBL Array
Initial States
For MIMO Case

Para Name
In/Out
Type
Description
stimulus
In
2D DBL Array
The stimulus signal into the system
response
In
2D DBL Array
The response output from the system
# of states
In
I32
Number of states
# p,f
In
I32
Estimation horizon
structure
In
Cluster
Information about the D matrix, K matrix and
selector
coefficients
initial state
Out
Cluster
The coefficients of state space model system
of state-
matrix, A, B, C, and D
space
4.
Out
2D DBL Array
Kalman gain
X0
Out
1D DBL Array
Initial States
Algorithm
A state space model in linear regression form can be represented as
qxt A B xt K
y C D u I et
t
t
Using the least squares criterion and you get the following equation to solve the parameter matrix:
qX t A B X t qX t T T
A B
C D min Y C D U Y X t ,U t
t t
t
Where
Xt T T
X t ,U t
U
t
(1)
denotes the pseudo inverse.
From the above equation, you can see that you can estimate the system matrix if you can estimate
the state sequence
Xt .
38
ni.com
A.
Estimation of System Matrix

The System Identification Toolkit defines the past and future Hankel data matrices in the following
way:
yt
qy
t
Yf

f 1
q yt
yt 1
qyt 1
f 1
q y y1
q 1 yt

q p y
Z p 1 t
q ut

p
q ut
q 1 yt 1
p
q yt 1
q 1ut 1
p
q ut 1
yt N 1
ut
, U qut
f
f 1
q f 1 yt N 1
q uu
ut 1
qut 1
f 1
q ut 1
ut N 1
f 1
q ut N 1
q 1 yt N 2 q 1 yt N 1
q p yt N 2 q p yt N 1
q 1ut N 2 q 1ut N 1
q p ut N 2 q p ut N 1
(2)
Where p is the past horizon, f is the future horizon. By iterating the system equations, you can get
the extended model, represented by the following equation:
Yf f X H f U f Gf E f
Where f is the extended system observability matrix.
0 0 0
D
C
CB
D 0 0
CA
, H CAB CB D 0
f
f

f 1

CA
f
2
CA B CB
0 0 0
I
CK
I
0 0
G f CAK CK I
0
2
CA K CK
0
0
0,
0
D
0
0
0
0
I
The Kalman state X is unknown, but that any Kalman state is estimated from past input and output
data is known, i.e.
U p
X Lu , Ly Lz Z p
Yp
Then you can have
Yf f Lz Z p H f U f Gf E f
Project out
Uf
by multiplying
U f
and remove noise
39
E f by multiplying Z Tp , you get
Yf U f Z Tp z Z p U f Z Tp
Where
z f Lz
and
p is the extended system controllability matrix. z
using the QR decomposition equivalents for
can be estimated
[U Tf , Z Tp ,Y fT ] .
R11 R12 R13

U , Z , Y Q1, Q2 , Q3 0 R22 R23
0 0 R33
T
T
z R23
R22 R22
R22
T
f
T
p
T
f
After obtaining the z , you can estimate
(3)
by performing SVD decomposition on
(4)
weighted
with Z p .
SVD
z Z p USV T
f U1
Because you already obtained the estimation of
Xt
by solving z Z p
z and f , then you can estimate the state sequence
f X t . That is
X t Tf f
And also the
With the
B.
qXt
(5)
pT z Z p
(6)
can be estimated similarly.
Yt U t X t and qXt
you can solve the equation 1 to obtain A, B, C, and D.
Estimation of Kalman Gain K
Yf Z G f E f
From
U f
T
R33
Q3T G f E f
You can get

Where
T
R33
Q3T comes from equation (3).
Transform the innovation process into unit variance white sequence, e1 , i.e.
e(t) Fe1 (t)
G f E f G f F E1 f Gf E1 f
T
Gf E1 f R33
Q3T
Because
Q3T
is an orthonormal matrix, you can choose E1 f
40
Q3T , then
ni.com
T
Gf R33
Then the Kalman gain K, actually KF, can be identified from the first block column of
as
G f , denoted
Gf 1 ,
F
CKF I
0 F
mm
Gf 1
0 1:m f 1,: KF
f 2
CA F
KF and F can be estimated using least squares, which then leads to
K KF F 1
C.
(7)
Estimation of Initial States X0

To write the equation
Yf f X H f U f Gf E f
at t = 0,
Y f 0 f X 0 H f U f 0 Gf E1 f 0
You can first compute Yf 0 for zero initial states to obtain
through the estimated system with X 0
By removing
H f U f 0 by processing the stimulus
0, , 0.
H f U f 0 from the double sides of the above equation you have:
G Y
1
f
f0
H f U f 0 Gf
f X 0 E1 f 0
(8)
You can estimate the X0 using least squares.
5.
Reference
[1]. Jin Wang and S. Joe Qin. A New Deterministic-Stochastic Subspace Identification Method
Using Principal Component Analysis. TWMCC Technical Report TWMCC-2004-01.
[2]. Peter Van Overschee and Bart De Moor. N4SID : Subspace Algorithms for the Identification of
Combined Deterministic-Stochastic Systems. Automatica, Special Issue on Statistical Signal
Processing & Control, Vol.30, No.1, pp.75-93, 1994.
[3]. Ljung, L. 1999. System Identification Theory for the User. 2nd ed. Prentice Hall.
si_Correlation Analysis.vi
1.
Introduction
This section includes the algorithms and references used by the si_Correlation Analysis.vi, which
is the core VI to estimate impulse responses of a system. The si_Correlation Analysis.vi is located
in the labview\vi.lib\addons\System Identification\Nonparametric Estimation Subs.llb directory.
41
2.
A discrete-time description of the system with an additive disturbance v(t ) can be represented as:
y(t ) g (k )u(t k ) v(t ) ,
t 0,1,2,
(1)
k 0
where u(t) and y(t) are the input and output of a system, and v(t) is the disturbance of the system.
g(k) is the impulse response of the system.
The purpose of this VI is to estimate the impulse response, g(k), of the system based on the input
and output of the system.
3.
Function API
Para Name
In/Out
Type
Description
stimulus signal
In
1D DBL Array
response
In
1D DBL Array
AR Order
In
I32
Order of the prewhitening filter
num of points
In
DBL
Length of impulse response sequence g(k)
Impulse
Out
1D DBL Array
Estimated impulse response
whiteness test
Out
XY Graph
Plot to indicated the effect of the prewhitening
rejected
Out
Boolean
Whether or not the estimated impulse response is
signal
Response
reliable
confidence
Out
DBL
Confidence level of the impulse response
level
4.
Algorithms [Page 169, Correlation Analysis, Ref 1]

According to Equation 1, the autocorrelation of the inputs,
input and output,
Ru , and cross-correlation between
Ryu , are related with
Ryu ( ) g (k )Ru (k )
(2)
k 0
If the input signal u(t) is white noise sequence, then
Ru ( ) E[u(t )u(t )] ( )
where
is the variance of u(t). Substituting Equation 3 to 1 yields

R ( )
g ( ) yu
(3)
(4)
Input Pre-Whitening
42
ni.com
If the input sequence is not exactly random white noise, large errors exist in the impulse response
calculated from the above correlation analysis method. A pre-whitening FIR filter is applied to
both the input and output sequences to whiten the input sequence. The coefficients of the prewhitening filter are exactly the coefficients of the AR model of the input sequence.
Whiteness Test
The stimulus signal after prewhitening is considered to be white noise. However, if the AR order is
not selected properly, this prewhitening might not be as good as expected. Therefore, you need to
test the whiteness of the prewhitened stimulus signal by computing and plotting the autocorrelation
and 99% confidence level with the System Identification Toolkit. If all Ru ( ),
are within
the confidence region, you have 99% confidence to say that the signal has been prewhitened.
( ) , where is the variance of the

~
~
signal. In other words, all Ru ( ), 0 are exactly zeros, while Ru (0) equals to the variance of
~
the white noise. Because of the limit length of the signal to compute Ru ( ) , each point of
~
Ru ( ), 0 is usually considered as zero mean Gaussian distribution with variance
Theoretically, the autocorrelation of a white noise is
2 / N [Page 512. Whiteness Test, Ref 1]. So the 99% confidence level of each point of
~
Ru ( ), 0 is computed by 2.576 ( 2.576 / N ) . The constant 2.576 is computed as
the following diagram:
5.
Reference
SI Estimate Impulse Response (Least Squares

Array).vi
1.
Introduction
This section includes the algorithms and references used by the SI Estimate Impulse Response
(Least Squares Array).vi. The SI Estimate Impulse Response (Least Squares Array).vi is the core
VI used to calculate an impulse response using the least squares method. The SI Estimate Impulse
Response (Least Squares Array).vi is located in the labview\vi.lib\addons\System
Identification\Nonparametric Estimation.llb directory.
43
2.
A discrete-time description of a system with an additive noise e(n) can be represented as:
yn
hk un k en ,
where u(n) and y(n) are the input and output of a system, and h(k) is the impulse response of the
system. The purpose of this VI is to estimate the impulse response, h(k), using the input and output
signals of the system.
3.
APIs
Para Name
In/Out
Type
Description
stimulus signal
In
1D DBL Array
response signal
In
1D DBL Array
impulse
In
Cluster
Number of points for which t<0 and t>=0,
response length
impulse
represented as
N t 0
and
N t 0
, respectively
Out
Cluster
The estimated impulse response
Out
Cluster
The standard deviation of the estimated
response
standard
deviation
4.
impulse response
Algorithms
Using Equation 1, you can calculate the following linear equations by truncating the impulse
response to a limited length.
AH E Y
(2)
where
h Nt 0
T ( p)
e( p Nt 0 )
h N 1
T
0
e( p 1 N )
( p 1)
t 0
, E
A
, H
, and
h( Nt 0 1)
(L 1)
e( L 1 Nt 0 )
hNt 0
y( p Nt 0 )
y ( p 1 N )
t 0
Y
y( L 1 Nt 0 )
(t ) u(t ) u(t 1)...u(t Nt 0 Nt0 1)T

p Nt 0 Nt 0 1
L is the length of the signals.
By finding the optimal H, you can minimize the following cost function:
44
ni.com
L1 Nt 0
2
e n
(3)
n p Nt 0
You then can deduce that the optimal solution for the above cost function is
H AT A AT Y
(4)
Equation 4 also is known as the least squares method. The covariance matrix of the estimated
impulse response is:
C e AT A
(5)
where C is the covariance matrix and e is the variance of the noise e(n). [Ref 1]
5.
References
[1] Ljung, L. 1999. System Identification Theory for the User. 2nd ed. Prentice Hall, Page 284,
Equation 9.18.
si_Frequency Analysis.vi
1.
Introduction
This section includes the algorithms and references used by the si_Frequency Analysis.vi, which is
the core VI to estimate frequency response of a system. The si_ Frequency Analysis.vi is located in
the labview\vi.lib\addons\System Identification\Nonparametric Estimation Subs.llb directory.
2.
The discrete-time description of a system can be represented as
Ryu ( ) g (k )Ru (k )
(1)
k 0
where
Ru
is the autocorrelation of the input of a system, and
Ryu
is the cross-correlation between
the input and output of a system. g(k) is the impulse response of the system. By using the Fourier
transform on both sides of Equation 1, you obtain
yu (e j ) G(e j )u (e j )
G(e j ) is known as the frequency response of the system.
(2)
The purpose of this VI is to estimate
the frequency response of a system based on its measured input and output signal. Refer to the
si_Correlation Analysis.vi for more information about the discrete-time description of a system.
3.
Function API
Para Name
stimulus signal
In/Out
In
Type
1D DBL Array
Description
45
response
In
1D DBL Array
In
I32
Order of the prewhitening filter
In
I32
Number of points of the frequency response
signal
Window
length
num of freq
bins
sequence
freq in log?
In
Boolean
Boolean to select whether the frequency points are

distributed linearly or logarithmically
amplitude in
In
Boolean
dB?
Boolean to determine whether the amplitude of the

frequency response is in dB unit
sampling rate
In
DBL
Amplitude
Out
XY Graph
Amplitude of the frequency response
Out
XY Graph
Phase of the frequency response
Out
Cluster
The standard deviation of the amplitude and phase
Response
Phase
Response
Std deviations
response
4.
Algorithms [Page 178, Spectral Analysis, Ref 1]

From Equation 2, the frequency response of the linear system can be computed by[Equation 6.51,
Ref 1]:
Nyu ()
G(e ) N
u ()
(3)
where
N 1
Nyu () 1 Ryu ( )e j
N 0
N 1
uN() 1 Ru ( )e j
N 0
(4)
(5)
However, due to the limit length of the signal to compute Equation 4 and 5, the frequency leakage
of the Fourier transform may corrupt the estimation of frequency response. So normally the
following window is applied to reduce the leakage of the Fourier transform.
yu (e j )
1 N 1
Ryu ( )w( )e j
N 0
u (e j )
w( )
1 N 1
Ru ( )w( )e j
N 0
(6)
(7)
is known as the lag window. The lag window controls the trade-off of the variance and bias
of the frequency response estimation. The narrower the lag window, which is applied on the
46
ni.com
correlation function in time domain, the more different frequencies will be weighted and then the
estimated frequency response is smooth. However, the estimation may be biased considerably.
Residual Spectrum
The spectrum of the residual of a system can be computed with [Equation 6.78, Ref 1]:
Nyu ()
N
N
v () y () N
u ()
(8)
Standard Deviation of Frequency Response

The standard deviation of the frequency response estimation can be estimated with [Equation 6.59,
Ref 1]:
E | G N (e j ) EG N (e j ) |2
where N is the signal length and
W ( ) 0.75
()
1
W ( ) v
N
u ()
(9)
because a Hanning window is used.
is the
window length. [Equation 6.57, Ref 1]

5.
Reference
SI Estimate Partially Known Continuous Transfer

Function Model (General SISO Array).vi
1.
Introduction
This section includes the algorithms and references used by the SI Estimate Partially Known
Continuous Transfer Function Model (General SISO Array).vi. The SI Estimate Partially Known
Continuous Transfer Function Model (General SISO Array).vi is the core VI used to estimate a
partially known state-space model. The SI Estimate Partially Known Continuous Transfer
Function Model (General SISO Array).vi is located in the labview\vi.lib\addons\System
Identification\Partially Known Model Estimation.llb directory.
2.
The partially known transfer function model is represented by the following two equations:
K p (1 a1s ... al sl )
1 T s1 2r s / s / w
i
p
j 2
1 T s1 2r s / s / w
i
p
(1)
sh K p (1 a1s ... al sl )
e sTd
j 2
e sTd
(2)
47
Where
K p is the static gain

Td is the time delay
Tpi is the first-order time constant
is the natural frequency
r j is the damping ratio

In Equation 1, h is the number of poles at the origin in the model:
h i 2 j n , where n
is the denominator order
l m , where m is the numerator order

In Equation 2, h is the number of zeros at the origin in the model:
i 2 j n , where n is the denominator order
h l m , where m is the numerator order

st
The special 1 order model is represented as:
K p sTd
e
1 Tp s
(3)
The special 2nd order model is represented as:
Kp
e sTd
1 2rs / s 2 / 2
3.
(4)
APIs
Para Name
In/Out
Type
Description
stimulus signal
In
1D DBL Array
response signal
In
1D DBL Array
system model in
In
Object containing the

partially known
information about a
system model
sampling rate
In
DBL
system model
Out
Object containing
out
model
coefficients of
Out
Cluster
transfer function
model
delay (s)
Out
Numeric
The estimated
den info
Out
Cluster
The estimated denominator
Td
information
48
ni.com
static gain
4.
Out
Numeric
The estimated
Kp
Algorithm
You can estimate a partially known continuous transfer function model by refining initial estimates
with Gauss-Newton optimization routines. The purpose of optimization is to minimize the
following cost function:
VN
where
1 N 1 2
1 N 1
e
(
k
,
yk, yk
N k 1 2
N k 1 2
(5)
yk , is the predicted response, and yk is the measured
response.
If you have prior knowledge about
K p , Td , Tpi , j , and r j ,
you can specify these variables
using the SI Create Partially Known Continuous Transfer Function Model VI. If you do not
specify these variables using the SI Create Partially Known Continuous Transfer Function Model
VI, the SI Estimate Partially Known Continuous Transfer Function Model VI can find the initial
estimates for these variables. Refer to the SI Estimate Continuous Transfer Function Model (SISO
Array).vi for more information about finding the initial estimates for the coefficients of a
continuous transfer function model.
SI Estimate Partially Known State-Space Model (MIMO

Array).vi
1.
Introduction
This section includes the algorithms and references used by the SI Estimate Partially Known StateSpace Model (MIMO Array).vi. The SI Estimate Partially Known State-Space Model (MIMO
Array).vi is the core VI used to estimate partially known state-space models. The SI Estimate
Partially Known State-Space Model (MIMO Array).vi is located in the
labview\vi.lib\addons\System Identification\Partially Known Model Estimation.llb directory.
2.
Discrete state-space models can be represented as:
xt 1 Axt But Ket

yt Cxt Dut et
Continuous state-space models can be represented as:
dxt
Axt But Ket
dt
49
yt Cxt Dut et
The goal of this VI is to estimate a user-defined state-space model.
3.
APIs
Para Name
In/Out
Type
Description
stimulus signal
In
2D DBL Array
response signal
In
2D DBL Array
system model in
In
Object containing the

partially known
information about a
system model
sampling rate
In
DBL
system model out
Out
Object containing
model
coefficients of
Out
Cluster
The estimated values for A, B, C, and
state-space model
variables
D.
Out
1D Cluster Array
The name and optimized value of

each variable
4.
Kalman gain
Out
2D DBL Array
The estimated K matrix
initial states
Out
1D DBL Array
The estimated initial states
Algorithm
You can estimate a partially known state-space model by refining initial estimates using Gauss-Newton
optimization routines. The purpose of optimization is to minimize the following cost function:
1 N 1
1 N 1
VN e2 (k , ) yk , yk
N k 1 2
N k 1 2
where
(1)
yk , is the predicted response, and yk is the measured
response.
You can specify the initial estimates for the variables of a model using the SI Create Partially
Known State-Space Model VI.
Recursive Model Estimation VIs

1.
Introduction
This section includes the algorithms and references used by the si_Recursive ARMAX Model
(SISO).vi, si_Recursive OE Model (SISO).vi, si_Recursive BJ Model (SISO).vi, and si_Recursive
50
ni.com
GL Model (MISO).vi. These VIs are located in the labview\vi.lib\addons\System

Identification\Online Estimation Subs.llb directory.
2.
Algorithms
Recursively parametric model estimation is generally divided into two steps. The first step is to
construct and update the state vector of a system. The second step is to optimize parameter values
by using different update methods.
The generalized model can be rewritten as
where
y(t | ) T (t, ) in a linear regression form.
T (n) is the data vector constructed from the past input-output data u(n) and y(n) .
is
the parametric vector with model parameters.

The RLS recursive schema is [11.12a, b, c, 1, page 365, Ref 1]
^
(t 1) L(t ) (t )
^
(t ) y(t ) y(t )
L(t )
P(t 1) (t )
(t ) (t )T P(t 1) (t )
1
P(t 1) (t ) (t )T P(t 1)
P(t )
P(t 1)
(t )
(t ) (t )T P(t 1) (t )
t
If you select
R(t ) (k ) T (k )
[10.40, 1, page 328, Ref 1], (k ) is a gradient vector [9.6,
k 1
page 281, Ref 1], and the recursive schema changes into
^
(t 1) L(t ) (t )
^
(t ) y(t ) y(t )
L(t )
P(t 1) (t )
(t ) (t )T P(t 1) (t )
P(t )
1
P(t 1) (t ) (t )T P(t 1)
P
(
t
1)
(t )
(t ) (t )T P(t 1) (t )
For different model structures, you need to compute different gradient vector (k )
si_Recursive ARMAX Model (SISO).vi

This VI is the core VI to recursively estimate the response of ARMAX (SISO) model based on the
input stimulus. The ARMAX (SISO) model is described as
A(q) y(n) B(q)u(n k ) C(q)e(n)
51
where A(q) 1
Na
a q
i 1
Nb
B(q) bj q
, C(q) 1
j 0
Nc
c q
i 1
and q is the shift operator,
which means
q i y(n) y(n i) .
The general ARMAX (SISO) can be reform as the following linear regressions structure
y(n) T (n)
where
T (n) is the data vector constructed from the past input-output data u(n) and y(n) :
(n) [ y(n 1),..., y(n Na ), u(n k ),..., u(n k N1,b ), (n 1),..., (n Nc )]T .
is the parametric vector with model parameters:
[a1,..., aNa , b0 ,..., bNb , c1,..., cNc ]T
Algorithm Reference
^
(t, ) is the gradient vector of y(t ) as
(t, ) [
^
^
^
^
^
^
y(t ),..,
y(t ),
y(t ),...,
y(t ),
y(t )
y(t )]
a1
aNa
b0
bNb
c1
cNc
where
C(q)
^
y(t ) qi y(t ), i 1,..., Na
ai
C(q)
^
y(t ) qiu(t ), i 0,..., Nb
bi
C(q)
^
y(t ) qi (t ), i 0,..., Nc
ci
According to the definition of regression vector (t , ) , you can have
C(q) (t, ai ) (t, ai ) , (t k, ai ) (t, aik )

That is
(t, ai ) (t, ai ) c1 (t 1, ai ),,, cN (t Nc , ai )

(t, ai ) c1 (t, ai 1 ),,, cN (t, ai N )
c
C(q) (t, bi ) (t, bi ) , (t k, bi ) (t, bik )

That is
(t, bi ) (t, bi ) c1 (t 1, bi ),,, cN (t Nc , bi )

(t, bi ) c1 (t, bi 1 ),,, cN (t, bi N )
c
52
ni.com
C(q) (t, ci ) (t, ci ) , (t k, ci ) (t, cik )

That is
(t, ci ) (t, ci ) c1 (t 1, ci ),,, cN (t Nc , ci )

(t, ci ) c1 (t, ci 1 ),,, cN (t, ci N )
c
si_Recursive OE Model (SISO).vi

This VI is the core VI to recursively estimate the response of OE (SISO) model based on the input
stimulus. An OE (SISO) model is described as
y(n)
where
B(q)
u(n k ) e(n)
F (q)
Nb
Nf
j 0
i 1
B(q) bj q j , F (q) 1 fi qi and q is the shift operator, which means

q i y(n) y(n i) .

The general OE(SISO) can be reform as the following linear regressions structure:
y(n) T (n)
where
(n) [u(n k ),..., u(n k Nb )1(n 1), w(n 1),..., w(n N f )]T
where
w(t ) f1w(t 1) f2 w(t 2) ... f N f w(t N f )

b0u(t k ) b1u(t k 1) ... b Nb u(t k Nb )
[b0 ,..., bNb , f1,..., f N f ]T
Algorithm Reference
^
(t, ) is the gradient vector of y(t ) defined as
(t, ) [
^
^
^
^
y(t ),...,
y(t ), y(t ),...,
y(t )]
b0
bNb
f1
f N f
where
F (q)
^
y(t ) qiu(t ), i 0,..., Nb
bi
53
F (q)
^
y(t ) qi w(t ), i 1,..., N f
fi
F (q) (t, bi ) (t, bi ) , (t k, bi ) (t, bik )

That is
(t, ai ) (t, bi ) f1 (t 1, bi ),,, f N (t N f , bi )

f
(t, bi ) f1 (t, bi 1 ),,, f N f (t, bi N f )

F (q) (t, fi ) (t, fi ) , (t k, fi ) (t, fi k )
That is
(t, fi ) (t, fi ) f1 (t 1, fi ),,, f N (t N f , fi )

f
(t, fi ) f1 (t, fi 1 ),,, f N f (t, fi N f )

si_Recursive BJ Model (SISO).vi
This VI is the core VI to recursively estimate the response of BJ (SISO) model based on the input
stimulus. A BJ (SISO) model is described as
y(n)
B(q)
C(q)
u(n k )
e(n)
F (q)
D(q)
where
Nb
Nf
Nc
Nd
j 0
i 1
i 1
i 1
B(q) bj q j , F (q) 1 fi qi , C(q) 1 ci qi , D(q) 1 di qi and
q is
the shift operator, which means
q i y(n) y(n i) .
The general OE(SISO) can be reformed as the following linear regressions structure:
y(n) T (n)
where
(n) [u(n k ),..., u(n k Nb )1(n 1), w(n 1),..., w(n N f ),

(n 1),..., (n Nc ), v(n 1),..., v(t Nd )]T
where
w(t ) f1w(t 1) f2 w(t 2) ... f N f w(t N f )

b0u(t k ) b1u(t k 1) ... b Nb u(t k Nb )
54
ni.com
v(t ) d1v(t 1) d2v(t 1) ... dnd v(t Nd )

e(t ) c1e(t ) c2e(t 2) ... cnc e(t Nc )
[b0 ,..., bNb , f1,..., f N f , c1,..., cNc , d1,..., dNd ]T

Algorithm Reference
^
(t, ) is the gradient vector of y(t ) defined as:
(t, ) [
^
^
^
^
^
^
^
^
y(t ),...,
y(t ), y(t ),...,
y(t ),
y(t ),...,
y(t ),
y(t ),...,
y(t )]
b0
bNb
f1
f N f
c1
cNc
d1
dNd
where
C(q)F (q)
^
y(t ) D(q)qiu(t ), i 0,..., Nb
bi
C(q)F (q)
^
y(t ) D(q)qi w(t ), i 1,..., N f
fi
C(q)
^
y(t ) qi (t ), i 1,..., Nc
ci
C(q)
^
y(t ) qi v(t ), i 0,..., Nd
di
According to the definition of regression vector
(t, ) , you can have
C(q)F (q) (t, bi ) D(q)(t, bi ) , (t k, bi ) (t, bik ) ,

If you define C(q) F (q) G(q) , that is
(t, bi ) (t, bi ) d1 (t 1, bi ),..., d N (t Nd , bi ) g1 (t 1, bi ),,, g N (t N g , bi )

d
(t, bi ) d1 (t, bi 1 ),..., d Nd (t, bi Nd ) g1 (t, bi 1 ),,, g Ng (t, bi Ng )

C(q)F (q) (t, fi ) D(q)(t, fi ) , (t k, fi ) (t, fi k )
If you define C(q) F (q) G(q) , that is
(t, fi ) (t, fi ) d1 (t 1, fi ),..., d N (t Nd , fi ) g1 (t 1, fi ),,, g N (t N g , fi )

d
(t, fi ) d1 (t, fi 1 ),..., d Nd (t, fi Nd ) g1 (t, fi 1 ),,, g Ng (t, fi Ng )

That is

(t, ci ) c1 (t, ci 1 ),,, cN (t, ci N )
c
55
C(q) (t, di ) (t, di ) , (t k, di ) (t, dik )

That is
(t, di ) (t, di ) c1 (t 1, di ),,, cN (t Nc , di )

(t, di ) c1 (t, di 1 ),,, cN (t, di N )
c
si_Recursive GL Model (MISO).vi

This VI is the core VI to recursively estimate the response of general linear (SISO) model based on
the input stimulus. A BJ (SISO) model is described as
nu
A(q) y(n)
i 1
Bi (q)
C(q)
ui (n k )
e(n)
Fi (q)
D(q)
Na
A(q) 1 ai q i
where
i 1
Bi (q)
Ni ,b 1
i, j
j 0
, C(q) 1
Nc
c q
i 1
, D(q) 1
Nd
d q
i 1
Ni , f
F (q) 1 fi, j q j and

j 1
q is the shift operator, which means
q i y(n) y(n i) .
Na , Ni,b , Nc , Nd , Ni, f are A order, B orders, C order, D order and F
orders of the general linear model.

The general linear model can be reformed into the following linear regressions structure:
y(n) T (n)
where
(n) [ y(n 1),..., y(n Na ),

u1 (n k ),..., u1 (n k N1,b ),..., unu (n k ),..., unu (n k Nnu ,b ),
w1 (n 1),..., w1 (n N1, f ),..., wnu (n 1),..., wnu (n Nnu , f ),
(n 1),..., (n Nc ), v(n 1),..., v(t Nd )]T

where
wi (t ) fi ,1wi (t 1) fi ,2 wi (t 1) ... fi ,n f 1 wi (t Ni , f )
bi,1ui (t ki ) bi,2ui (t ki 1) ... bi,nb1 ui (t ki Ni,b 1)

56
ni.com
[a1,..., aNa , b1,0 ,..., b1, N1,b ,..., bnu ,0 ,..., bnu , Nnu ,b , f1,1,..., f1, N1, f ,..., fnu ,1,..., fnu , Nnu , f , c1,..., cNc , d1 ,..., dNd ]T
Algorithm Reference
^
(t, ) is the gradient vector of y(t ) as:
(t, ) [
^
^
y(t ),...,
y(t ),
a1
aNa
^
^
^
^
y(t ),...,
y(t ),...,
y(t ),...,
y(t ),
b1,0
b1, Nb
bnu ,0
bnu , Nnu ,b
^
^
^
^
y(t ),...,
y(t ),...,
y(t ),...,
y(t ),
f1,1
f1, N f
fnu ,1
fnu , Nnu , f
^
^
^
^
y(t ),...,
y(t ),
y(t ),...,
y(t )]
c1
cNc
d1
d Nd
where
C(q)
^
y(t ) D(q)qi y(t ), i 0,..., Na
ai
C(q)F (q)
^
y(t ) D(q)q jui (t ), i 1,..., nu , j 0,..., Ni,b
bi, j
C(q)F (q)
^
y(t ) D(q)q j wi (t ), i 1,..., nu , j 1,..., Ni, f
fi, j
C(q)
^
y(t ) qi (t ), i 1,..., Nc
ci
C(q)
^
y(t ) qi v(t ), i 1,..., Nd
di
C(q)F (q) (t, ak ) D(q)(t, ak ) , (t k, ai ) (t, aik ) .If defines

C(q)F (q) G(q)
That is
(t, ai ) (t, ai ) d1(t 1, ai ),..., d N (t Nd , ai ) g1 (t 1, ai ),,, g N (t N g , ai )

d
(t, ai ) d1 (t, ai 1 ),..., d Nd (t, ai Nd ) g1 (t, ai 1 ),,, g Ng (t, ai Ng )
57
C(q)Fi (q) (t, bi, j ) D(q)(t, bi, j ) , (t k, bi, j ) (t, bi, j k ) .If defines
C(q)Fi (q) Gi (q)
That is
(t, bi, j ) (t, bi , j ) d1 (t 1, bi , j ),..., d N (t Nd , bi , j ) gi ,1 (t 1, bi , j ),,, gi ,N (t N g , bi , j )

d
(t, bi, j ) d1 (t, bi, j 1 ),..., d Nd (t, bi, j Nd ) gi ,1 (t , bi , j 1 ),,, gi , Ng (t, bi , j Ng )
C(q)Fi (q) (t, fi, j ) D(q)(t, fi, j ) , (t k, bi, j ) (t, bi, j k ) . If defines
C(q)Fi (q) Gi (q)
That is
(t, fi, j ) (t, fi, j ) d1 (t 1, fi, j ),..., d N (t Nd , fi , j ) gn,1 (t 1, fi , j ),,, gn, N (t N g , fi , j )

d
(t, fi , j ) d1 (t, fi , j 1 ),..., dNd (t, fi , j Nd ) gn,1 (t , fi , j 1 ),,, gn, Ng (t, fi , j Ng )

That is

(t, ci ) c1 (t, ci 1 ),,, cN (t, ci N )
c
C(q) (t, di ) (t, di ) , (t k, di ) (t, dik )

That is
(t, di ) (t, di ) c1 (t 1, di ),,, cN (t Nc , di )

(t, di ) c1 (t, di 1 ),,, cN (t, di N )
c
3.
Reference
SI Model Simulation VIs

1.
Introduction
This section includes the algorithms and references used by the si_GL Model Sim (MISO).vi,
si_Std Response To Model.vi, si_ARX Model Sim (MIMO).vi, si_State-Space (MIMO) core.vi
and si_Transfer Function Sim Core.vi, which are the core VIs to simulate defined models in SI
toolkit. These VIs are located in the labview\vi.lib\addons\System Identification\Model Analysis
Subs.llb directory.
58
ni.com
2. Algorithms
The following are the algorithms for the SI Model Simulation VIs.
si_GL Model Sim (MISO).vi

This VI is the core VI to simulate general linear (MISO) model based on the input stimulus. The
general linear (MISO) model is described as
nu
A(q) y(n)
i 1
Bi (q)
C(q)
ui (n k )
e(n)
Fi (q)
D(q)
Na
A(q) 1 ai q i
where
i 1
Bi (q)
Ni ,b 1
Nc
Nd
Ni , f
j 0
i 1
i 1
j 1
bi, j q j , C(q) 1 ci qi , D(q) 1 di qi F (q) 1 fi, j q j and
q is the shift operator, which means
q i y(n) y(n i) .

The general linear model can be reformed into the following linear regressions structure:
y(n) T (n)
where
(n) [ y(n 1),..., y(n Na ),

u1 (n k ),..., u1 (n k N1,b ),..., unu (n k ),..., unu (n k Nnu ,b ),
w1 (n 1),..., w1 (n N1, f ),..., wnu (n 1),..., wnu (n Nnu , f ),
(n 1),..., (n Nc ), v(n 1),..., v(t Nd )]T

where
wi (t ) fi ,1wi (t 1) fi ,2 wi (t 1) ... fi ,n f 1 wi (t Ni , f )
bi,1ui (t ki ) bi,2ui (t ki 1) ... bi,nb1 ui (t ki Ni,b 1)

[a1,..., aNa , b1,0 ,..., b1, N1,b ,..., bnu ,0 ,..., bnu , Nnu ,b , f1,1,..., f1, N1, f ,..., fnu ,1,..., fnu , Nnu , f , c1,..., cNc , d1 ,..., dNd ]T
Algorithm Reference
59
The simulation of general linear model is to compute response
y(n) with stimulus u(t ) and noise
e(t ) as inputs. The standard deviation of response is also to be computed.

Response computation
From the definition of general linear model, the response
y(n) can be computed using the
following steps:
1) IIR filtering stimulus
coefficients of
F (q) being backward coefficients.
2) IIR filtering noise
D(q)
u(t ) with coefficients of B(q) being forward coefficients and
e(t ) with coefficients of C(q) being forward coefficients and coefficients of
being backward coefficients.
3) Summarizing the results of filtered stimulus
u(n) and filtered noise e(n) .
4) Filtering the summed result with coefficients of
A(q) being backward coefficients and 1 being
forward coefficients.
Note: the noise
e(n) must be rescaled by the inputted noise variance.
Standard deviation of response computation

The Standard deviation of response computation s implemented in the si_Std Response to
Model.vi
Lemma: let the
be P . If
close to
d -dimension vector
be an estimate with mean
0 , and its covariance matrix

^
f (0 ) is the p -dimensional random variable, asymptotically as
0 , you can have
becomes sufficiently
f (0) with good accuracy, defined as

^
f ( ) f (0 ) f ' (0 )( 0 )
where
f ' is the p d derivative matrix of f

^
with respect to
^
. This means
Covf ( ) E( f ( ) Ef ( ))( f ( ) Ef ( ))T f ' (0 )Pf ' (0 )T

The above equation is known as the Gauss approximation formula. [page 292, 9-56, Ref 1]
You therefore can define simulated response function to model parameters as
f (n, ) y(n, )
The covariance matrix of simulated response can be obtained according to the lemma:
Covy(n, ) y' (0 )P y' (0 )T
60
ni.com
P is the covariance matrix of model parameter vector . Thus, you need to compute the
derivative matrix of
y' (n, ) .
Matrix y (n, ) computation

'
1 j
y(n)
q y(n) q j ya (n), j 1,..., Na
a j
A(q)
1
y(n)
q jui (n k ) q jubi (n k ), j 0,..., Ni,b
b(i, j )
A(q)Fi (q)
Bi (q)
y(n)
q jui (n k ) q ju fi (n k ), j 1,..., Ni, f
f(i, j )
A(q)Fi (q)Fi (q)
1
y(n)
q j e(n) q j ec (n), j 0,..., Nc
c j
A(q)D(q)
C(q)
y(n)
q j e(n) q j e f (n), j 0,..., Nd
d j
A(q)D(q)D(q)
ya (n) can be obtained by IIR filtering y(n) with -1 being forward coefficients and coefficient of
A(q) being backward coefficients.
ubi (n) can be obtained by IIR filtering ui (n) with 1 being forward coefficients and coefficient of
A(q)Fi (q) being backward coefficients.
u f i (n) can be obtained by IIR filtering ui (n) with Bi (q)

coefficient of
being forward coefficients and
A(q)Fi (q)Fi (q) being backward coefficients.
ec (n) can be obtained by IIR filtering e(n) with 1 being forward coefficients and coefficient of
A(q)D(q) being backward coefficients.
e f (n) can be obtained by IIR filtering e(n) with C(q)

coefficient of
A(q)D(q)D(q) being backward coefficients.
Define the Vector
61
yvec' (n, ) [q1 ya (n),..., q Na ya (n), ub1 (n k ),..., q N1,b ub1 (n k ),..., ubnu (n k ),..., q
q1u f 1 (n),..., q
N1, f
u f 1 (n),..., q1u f nu (n),..., q
q1ec (n),..., q Nc ec (n), q1e f (n),..., q
N f
Nnu , f
Nnu ,b
ubnu (n k ),...,
u f nu (n),
e f (n)]
[ ya (n 1),..., ya (n Na ), ub1 (n k ),..., ub1 (n k N1,b ),..., ubnu (n k ),..., ubnu (n k Nnu ,b ),...,
u f 1 (n k 1),..., u f 1 (n k N1, f ),..., u f nu (n k 1),..., u f nu (n k Nnu , f ),
ec (n 1),..., ec (n Nc ), e f (n 1),..., e f (n N f )]
The derivative matrix therefore is defined as
yvec' (1, )
y' (t, )
yvec' ( N , )
The covariance matrix of simulated response
Covy(t, )
can be obtained using the Gauss
approximation formula.
The diagonal values of
Covy(t, ) is the variances of simulated response points. The standard
deviation is the square root of the variance.

si_ARX Model Sim (MIMO).vi
This VI is the core VI to compute response of ARX (MIMO) model based on the input stimulus.
The ARX model is described by the backward shift operator
q 1
as:
A(q) y(t ) B(q)u(t k ) e(t )

where
u(t )
and
y(t )
are the inputs and outputs of the system, respectively. k is the delay of the
input. e(t) is the disturbance of the system.

The time domain equation of an ARX (MIMO) model is described as
a
y1 (t ) a111y1 (t 1) a112y1 (t 2) ... a11n11a y1 (t n11
)
a
a121y2 (t 1) a122y2 (t 2) ... a12n12a y2 (t n12
)
...
a1N y 1 yN y (t 1) a1N y 2 yN y (t 2) ... a1N y n1aN yN y (t n1aN y )
y
b
b111u1 (t k11) b112u1 (t k11 1) ... b11n11b u1 (t k11 n11
1)
b
1)
...
b1Nu 1uNu (t k1Nu ) b1Nu 2uNu (t k1Nu 1) ... b1Nu n1bN uNu (t k1Nu n1bNu 1)
u
e1 (t )
62
ni.com
a
y2 (t ) a221y2 (t 1) a222y2 (t 2) ... a22n22a y2 (t n22
)
a
a211y1 (t 1) a212y1 (t 2) ... a21n21a y1 (t n21
)
...
a2 N y 1 yN y (t 1) a2 N y 2 yN y (t 2) ... a2 N y n2aN yN y (t n2aN y )
y
b
1)
b
1)
...
b2 Nu 1uNu (t k2 Nu ) b2 Nu 2uNu (t k2 Nu 1) ... b2 Nu n2b N uNu (t k2 Nu n2bNu 1)
u
e2 (t )
yN y (t ) aN y N y 1 yN y (t 1) aN y N y 2 yN y (t 2) ... aN y N y nNa N yN y (t nNa y N y )

y y
aN y 11y1 (t 1) aN y 12 y1 (t 2) ... aN y 1nNa 1 y1 (t nNa y 1 )

y
...
aN y ( N y 1)1 y( N y 1) (t 1) aN y ( N y 1) 2 y( N y 1) (t 2) ... aN y ( N y 1) nNa
y ( N y 1)
y( N y 1) (t nNa y ( N y 1) )
bN y 11u1 (t kN y 1 ) bN y 12u1 (t kN y 1 1) ... bN y 1n21b u1 (t kN y 1 nNb y 1 1)

bN y 21u2 (t kN y 2 ) bN y 22u2 (t kN y 2 1) ... bN y 2n12b u2 (t kN y 2 nNb y 2 1)
...
bN y Nu 1uNu (t kN y Nu ) bN y Nu 2uNu (t kN y Nu 1) ... bN y Nu n2b N uNu (t kN y Nu nNb y Nu 1)
u
eN y (t )
From the model definition, the response of
u1(t),..., uNu (t ) and noise e1 (t ),..., eNy (t )
y1 (t ),..., yN y (t ) , coefficients of the model, stimulus

form a recursive relationship. The response of the
model is computed in the point-by-point style in this VI.
si_State-Space (MIMO) core.vi

This VI is the core VI to compute response of state-space (MIMO) model based on the input
stimulus. A state-space (MIMO) model is described as
X (t 1) AX (t ) BU (t ) KE(t )
,
Y (t ) CX (t ) DU (t ) E(t )
where
U (t )
and
Y (t )
are the inputs and outputs of the system, respectively.
X (t ) is the state
variables.
63
For MIMO System
u1 (t )
u (t )
2
, t 1, 2,..., N , n is the number of inputs, N is the
U (t ) is input vector. U (t )
u

unu (t )
sample number.
y1 (t )
y (t )
2
, t 1, 2,..., N , n is the number of outputs, N is the
Y (t ) is output vector. Y (t )
y

yny (t )
sample number.
E(t ) is noise disturbance vector.
X (t ) is state vector.
e1 (t )
e (t )
2
, t 1,2,..., N
E(t )

eny (t )
x1 (t )
x (t )
2
, t 1, 2,..., N , n is the number of states
X (t )
x

xnx (t )
A is nx nx state transmission matrix. B is nx nu input gain matrix. C is ny nx output gain

matrix.
D is ny nu direct-input gain matrix. K is nx ny noise gain matrix.
Y (t ) , states X (t ) , stimulus U (t ) and noise
E(t ) form a recursive relationship. The response of the model is computed in the point-by-point
style in this VI.
si_Transfer Function Sim Core.vi
This VI is the core VI to compute response of a transfer function (MIMO) model based on the
input stimulus. The transfer function (MIMO) model is described as
Y (t ) G( z)U (t )
G( z) is polynomial fraction matrix, defined as
g11 ( z) g1nx ( z)
G( z)

g ( z) g ( z)
ny nx
ny 1
Where
64
ni.com
gij (z) is polynomial fraction in delay operator z1 :

gij ( z)
n z n n1z n1 ... 1z1 0

m z m m1z m1 ... 1z1 0
i indicates ith output in MIMO system.
j indicates jth
input in MIMO system.
u1 (t )
u (t )
2
U (t ) is input vector. U (t )
u

unu (t )
sample number.
y1 (t )
y (t )
2
, t 1, 2,..., N , n is the number of outputs, N is the
Y (t ) is output vector. Y (t )
y

yny (t )
sample number.
From the definition of the model, the computation of multiple outputs
Y (t )
can be carried out one
output by one output.
yi (t ) gij ( z)u j (t ), i 1,.., N y , j 1,.., Nu

The computation of
gij (z)u j (t ) can be carried out by IIR filtering u j (t ) with nominator of
gij (z) being forward coefficients and denominator of gij (z) being backward coefficients.
3.
Reference
SI Model Prediction VIs

1.
Introduction
This section includes the algorithms and references used by the si_GL Model Predict (MISO).vi,
si_SS Model predict vi, and si_SS State Response.vi, which are the core VIs to predict the
response of defined models in SI toolkit. These VIs are located in the
labview\vi.lib\addons\System Identification\Model Analysis Subs.llb directory.
2.
Algorithms
The following are the algorithms for the SI Model Prediction VIs.
65
si_GL Model Predict (MISO).vi

This VI is the core VI to predict the response of general linear (MISO) model based on the input
stimulus. A generalized system model is defined as
y(t ) G(q)u(t ) H (q)e(t )

K-Step Prediction
The k-step-ahead prediction of
y(t ) is deduced as [3.31, page 71, Ref 1]
y(t | t k ) Wk (q)G(q)u(t ) [1Wk (q)]y(t )

where
Wk (q) H k (q)H 1 (q) . [3.27, page 70, Ref 1]
A general linear (MISO) model is described as

nu
A(q) y(t )
i 1
Bi (q)
C(q)
ui (t k )
e(t )
Fi (q)
D(q)
where
Na
A(q) 1 ai q i
Bi (q)
i 1
Ni ,b 1
bi, j q j
Nc
j 0
C(q) 1 ci qi
i 1
Nd
D(q) 1 di qi
i 1
Ni , f
F (q) 1 fi, j q j and q is the shift operator, which means:

j 1
q i y(n) y(n i) .
Algorithm Reference
According to the definition of the k-step-ahead prediction of
y(t ) and general linear model
(MISO), you can have
Gi (q)
Bi (q)
C(q)
, i 1,..., Nu , H (q)
A(q)Fi (q)
A(q)D(q)
The k-step-ahead prediction of
y(t ) can be reformed as
C(q)
F (q)D(q)
y(t | t k )
u (t )
F (q)C(q) F (q)
^
nu
i 1
H k (q)Bi (q)
j 1,..,nu
j 1,..,nu
j 1,..,nu
66
Fi (q) H k (q) A(q)D(q)

C(q)
j 1,..,nu
F (q)
j 1,..,nu
Fi (q)
ni.com
y(t )
F (q)
(q)
g (q ), n 1,..., k
A(q) D(q) F (q)
C(q)
H k (q) can be computed by H k
j 1,..,nu
j 1,..,nu
g (qn ), n 1,..., k can be calculated by using the impulse response of
an IIR filter.
F ( q)
with coefficients of C(q) F (q) being forward coefficients and
A(q) D(q) F (q)
C ( q)
j 1,..,nu
j 1,..,nu
A(q)D(q)
j 1,..,nu
F (q) being backward coefficients.
j 1,..,nu
So the k-step-ahead prediction of
H k (q)Bi (q)
Fi (q)C(q)
C(q)
being forward coefficients and coefficients of
F (q)
j 1,..,nu
y(t ) can be computed by IIR filtering ui (t ) with coefficients of
Fj (q)D(q)
j 1,..,nu
j 1,..,nu
of
being backward coefficients, and by IIR filtering
Fi (q) H k (q) A(q)D(q)
C(q)
F (q)
j 1,..,nu
coefficients of
j 1,..,nu
y(t ) with coefficients
Fi (q)
being backward coefficients.
si_SS Model Predict.vi

This VI is the core VI to predict the response of state-space (MIMO) model based on the input
stimulus. A state-space (MIMO) model is described as
X (t 1) AX (t ) BU (t ) KE(t )
,
Y (t ) CX (t ) DU (t ) E(t )
where
U (t )
and
Y (t )
are the inputs and outputs of the system, respectively.
X (t ) is the state
variables.
For MIMO System
u1 (t )
u (t )
2
U (t ) is the input vector. U (t )
u

unu (t )
sample number.
67
y1 (t )
y (t )
2
, t 1, 2,..., N , n is the number of outputs, N is
Y (t ) is the output vector. Y (t )
y

yny (t )
the sample number.
E(t ) is the noise disturbance vector.
X (t ) is the state vector.
e1 (t )
e (t )
2
, t 1,2,..., N
E(t )

eny (t )
x1 (t )
x (t )
2
, t 1, 2,..., N , n is the number of states
X (t )
x

xnx (t )
A is nx nx state transmission matrix. B is nx nu input gain matrix. C is ny nx output gain

matrix.
D is ny nu direct-input gain matrix. K is nx ny noise gain matrix.
Y (t ) , states X (t ) , stimulus U (t ) and noise
E(t ) form a recursive relationship. The response of the model is computed in the point-by-point
style in this VI.
Algorithm Reference
One-step Prediction
The conditional expectation of
^
y(t ) is the Gaussian process, given by [4.86, page 98, Ref 1]
X (t 1) A X (t ) BU (t ) K[Y (t ) C X (t ) DU (t )]
^
Y (t ) C X (t ) DU (t )
This can be reformed as
^
^
U (t )
X (t 1) [ A KC] X (t ) [ B KD, K ][
]
Y (t )
^
Y (t ) C X (t ) DU (t )
From
U (t ), Y (t ) , the one-step prediction of Y (t ) can be deduced from the above equation.
K-Step Prediction [10.99, page 345, Ref 1]
68
ni.com
The K-step prediction is also to find the formula between the predicted
Y (t ) and
U (t ), Y (t ) as the one-step prediction. In the calculation of Y (t ) , the formula only uses

outputs Y (t ) up to time t k .
From the state-space model equation, you can find
Y (t k ) CX (t k ) DU (t k ) E(t k )
CAX (t k 1) CBU (t k 1) CBKE(t k 1)
DU (t k ) E(t k )
...
CAk X (t ) CAk 1BU (t ) CAk 2 BU (t 1) ...
CBU (t k 1) DU (t k )
CAk 1KE(t ) CAk 2 KE(t 1) ...
CKE(t k 1) E(t k )
The state space model can be rewritten as the matrix multiplications, as shown by the
following equation:
Y (t k ) CAk X [CAk 1B, CAk 2 B,..., D][U (t ),U (t 1),...,U (t k )]T V

where
V is the noise influence matrix.
By discarding the noise influence matrix
V , the vector Y of k-step predictor is defined as
Y (t k ) CAk X (t ) [CAk 1B, CAk 2 B,..., D][U (t ),U (t 1),...,U (t k )]T

The formula can be reformed as
Y^ (1)
T
X (1 k )
U (1 k ),U (1 k 1),...,U (1)
X (2 k )
^ ^
[CAk 1B, CAk 2 B,..., D] U (2 k ),U (2 k 1),...,U (2)
Y (1) Y (t k ) CAk
X ( N k )
U ( N k ),U ( N k 1),...,U ( N )
Y ( N )
U (1 k ),U (1 k 1),...,U (1)

U (2 k ),U (2 k 1),...,U (2)
k 1
k 2
[CA B, CA B,..., D]
U ( N k ),U ( N k 1),...,U ( N )
filtering
can be computed by FIR
[U (1),U (2),...,U ( N )] with coefficients of [CAk 1B, CAk 2 B,..., D] .
si_SS State Response.vi

This VI is a subVI to compute state response of a state-space (MIMO) model based on the input
stimulus.
69
The formula of state response is defined as
X (t 1) AX (t ) BU (t )
X (t ) can be computed by a recursive process.
3.
Reference
si_Residual Analysis.vi
1.
Introduction
This section includes algorithms and references used by the si_Residual Analysis.vi, which is the
core VI to calculate the autocorrelation of prediction error and cross-correlation between
prediction error and stimulus signal. The si_Residual Analysis.vi is located in the
labview\vi.lib\addons\System Identification\Model Analysis Subs.llb directory.
2.
You can use this VI to calculate the autocorrelation of prediction error and cross-correlation
among the prediction error, stimulus signal, and their confidence interval.
3.
APIs
Para Name
prediction
In/Out
Type
Description
In
1D CDB Array
Prediction error
input signal u
In
2D DBL Array
Input signal
cov(e)
Out
1D CDB Array
Normalized auto-correlation function of the
error
prediction errors
confidence of
Out
DBL
Confidence of
Out
2D DBL Array
Normalized cross-correlation function of the
cov(e)
cov(e)
cov(u, e)
prediction errors and input signal

confidence of
Out
1D CDB Array
Confidence of
cov(u, e)
cov(u, e)
4.
Algorithms [Page 511-513, Ref.1]
cove ( )
The auto-correlation of prediction error is defined as
ReN ( )
1 N
e(t)e(t )
N t 1
70
(1)
ni.com
where N is the signal length. The normalized auto-correlation of prediction error is
cove ( )
where
cove ( )
N ReN ( ) is a zero mean Gaussian distribution with variance 2 [P.512, Ref 1],
ReN ( ) , is a zero mean Gaussian distribution
the auto-correlation of prediction error,

with variance
2 / N .
So the confidence interval of the
Na
where
(2)
1 N 2
e (t) ReN (0) , which is the variance of the prediction error.
N t 1
Confidence of
Since
ReN ( )
denotes the
ReN ( )
is
(3)
level of the normalized Gaussian distribution , N (0,1) , and
0.99% in this problem. Then the confidence interval of cove ( )

Na
is
1
N
(4)
coveu ( )
The cross-correlation between the prediction error, e(t), and the input signal, u(t), is
defined as
ReuN ( )
1 N
e(t)u(t )
N t 1
(5)
where N is the signal length. The normalized cross-correlation of prediction error is
coveu ( )
N
e
N
u
R (0)R (0)
Confidence of
Since
ReuN ( )
(6)
coveu ( )
N ReuN ( ) is a zero mean Gaussian distribution with variance P1 [Eq. (16.64),
P.513, Ref 1], where
P1
k N
The auto-correlation of prediction error,

with variance
P1 / N .
(k ) RuN (k )
And then the confidence interval of
(7)
ReuN ( ) , is a zero mean Gaussian distribution
So the confidence interval of
Na
N
e
P1
N
ReuN ( ) is
(8)
coveu ( )
71
is
1
ReN (0)RuN (0)
5.
Na
P1
N
(9)
Reference
si_Convert to Transfer Function Model.vi

1.
Introduction
This section includes the algorithms and references used by the si_Convert to Transfer Function
Model.vi, which is the core VI to convert defined models into transfer function models. The
si_Convert to Transfer Function Model.vi is located in the labview\vi.lib\addons\System
Identification\Model Conversion Subs.llb directory.
2.
You can use this VI to convert an AR, ARX (SISO, MISO, and MIMO), ARMAX (SISO, MISO),
OE (SISO, MISO), BJ (SISO, MISO), general linear model (SISO, MISO), or state-space model
(SISO, MISO, and MIMO) into the corresponding transfer function model.
3.
APIs
Para Name
system model in
In/Out
In
Type
Description
SI object
Contains the information of the model to be

converted.
4.
realization type
In
Text ring
Determines the type of state-space realization.
tolerance
In
DBL
Determines the zero-pole cancellation.
system model out
Out
SI object
Contains the information of the model.
Algorithms
Transfer Function Model Equivalents of General Linear Models
This algorithm is used by the si_GL2TF (MISO).vi.
If the transfer function between i-th input and output of a MISO general linear model is
Bi (q)
a1q1 amqm
Gi (q)
A(q)Fi (q) 1 b1q1 bmqm

It gives equivalent transfer function model to multiply
Gi (q)
(1)
q m on the nominator and denominator of
as
Gi (q)
a1qm1 qm
qm b1qm1 bm
72
(2)
ni.com
The transfer function between the noise and output of a MISO general linear model is
H (q)
C(q)
D(q)
(3)
Transfer Function Model Equivalents of State-Space Models [P. 651, Ref 1]

This algorithm is used by the si_SS2TF Matrix Combine.vi.
The state-space representation of a given MIMO system is
x(t 1) Ax(t ) Bu(t ) Ke(t )

y(t ) Cx(t ) Du(t ) e(t )
(4)
Performing Z transform on the state-space equations yields
( zI ) X ( z) AX ( z) BU ( z) KE( z)
Y ( z) CX ( z) DU ( z) E( z)
(5)
So
X ( z) ( zI A)1[BU ( z) KE( z)]

Then
G( z)
Y ( z)
C( zI A)1 B D
U ( z)
(6)
H ( z)
Y ( z)
C( zI A)1 K I
E( z )
(7)
Further [Ref 1]
where
C( zI A)1 B
and
det( zI A BC) det( zI A)

det( zI A)
det( zI A) is the characteristic polynomial of matrix A .
The equivalent transfer function representation for the determinant part can be presented as
G( z) C( zI A)1 B D
det( zI A BC) det( zI A)(D 1)

det( zI A)
(8)
Each column vector of

column by column of
B is for one input. Each row vector of C is for one output. By selecting
B and row by row of C , MIMO system of G( z) can be computed channel
by channel.
The equivalent transfer function representation of the noise part is
H ( z) C( zI A)1 K I
5.
det( zI A KC) det( zI A)(I 1)

det( zI A)
(9)
Reference
73
[1] T. Kailath, Linear Systems, Prentice Hall, Prentice Hall, 1980.
si_Convert to State Space Model.vi

1.
Introduction
This section includes the algorithms and references used by the si_Convert to State-Space
Model.vi, which is the core VIs to convert defined models into state-space models. The si_Convert
to State-Space Model.vi is located in the labview\vi.lib\addons\System Identification\Model
Conversion Subs.llb directory.
2.
You can use this VI is to convert an AR, ARX (SISO, MISO, and MIMO), ARMAX (SISO,
MISO), OE (SISO, MISO), BJ (SISO, MISO), general linear model (SISO, MISO), transfer
function (SISO, MISO and MIMO), or pole-zero model (SISO, MISO and MIMO) into the
equivalent state-space model.
3.
APIs
Para Name
system model in
In/Out
In
Type
Description
SI object

converted.
4.
realization type
In
Text ring
tolerance
In
DBL
system model out
Out
SI object
Algorithms
State-Space Model Equivalents of Transfer Function Models
This algorithm is used by the si_TF2SS Realization (MIMO).vi
The transfer function representation of a given MIMO system is defined by the following
polynomial fraction matrix:
g11 (q) g1nu (q)
G(q)

g (q) g (q)
ny nu
ny 1
Where
gij (q) is the polynomial fraction in delay operator q1 .

gij ( z)
an qn an1qn1 ... a1q1 a0

bmqm bm1qm1 ... b1q1 b0
i indicates the ith
output in MIMO system.
74
ni.com
j indicates the jth
input in MIMO system.
G( z) can be reformed as
G(q)
where
1
An qn An1qn1 ... Aq
A0
1
m
m1
1
Bmq Bm1q ... B1q B0
A0 ,..., An , B0 ,..., Bm is ny nu matrix.
The equivalent observable state-space realization of the transfer function model can be performed
as
Am1 1 0 ... 0
A
0 1 ... 0
A m2 ...
A0 0 0 ... 0
Bm1 Am1Bm
B A B
B m2 m2 m
B0 A0 Bm
C 1
D Bm
0 0
0 0
State-Space Model Equivalents of ARX Model (MIMO)

This algorithm is used by the si_ARX To SS (MIMO).vi.
According to the definition of ARX model (MIMO), an ARX model (MIMO) can be reformed as
Ey(t) + A1 y(t - 1) + ... + ANa y(t - Na ) = B0u(t - k ) + ... + BNb u(t - k - Nb ) + Ee(t )
Where
y(t ) is output vector [ y1 (t), y2 (t),..., yny (t)]T , u(t ) is input vector
[u1 (t ), u2 (t ),..., unu (t )]T , E is identity matrix, A1,..., ANa is ny ny matrix, B0 ,..., BNb is
ny nu matrix. k is the input delay.

The deterministic part of an ARX (MIMO) model can be further reformed into a transfer function
(MIMO) representation, which then can be further converted into a state-space model by
observable state-space realization.
B0q- k + ... + BNb q- k- Nb

B0q Na - k + ... + BNb q Na - k- Nb
G(q) =
=
E + A1q- 1 + ... + ANa q- Na
E Na + A1q Na - 1 + ... + ANa
For the noise part, the same conversion method is used as that of the determinant part.
5.
Reference
75
si_Convert to Pole-Zero Model.vi

1.
Introduction
This section includes the algorithms and references used by the si_Convert to Pole-Zero Model.vi,
which is the core VI to convert defined models into pole-zero models. the si_Convert to Pole-Zero
Model.vi is located in the labview\vi.lib\addons\System Identification\Model Conversion Subs.llb
directory.
2.
You can use this VI is to convert an AR, ARX (SISO, MISO, and MIMO), ARMAX (SISO,
MISO), OE (SISO, MISO), BJ (SISO, MISO), general linear model (SISO, MISO), state-space
model (SISO, MISO, and MIMO), or transfer function model (SISO, MISO, and MIMO) into the
equivalent pole-zero model.
3.
APIs
Para Name
In/Out
system model in
In
Type
Description
SI object

converted.
4.
realization type
In
Text ring
tolerance
In
DBL
system model out
Out
SI object
Algorithms
This algorithm is used by the si_TF2ZP (MIMO).vi and the si_TF2ZP plus STD (MIMO).vi.
The pole-zero model representations of a given MIMO system is defined by the following
polynomial fraction matrix:
g11 (q) g1nu (q)
G(q)

g (q) g (q)
ny nu
ny 1
where
gij (q) is the polynomial fraction in delay operator q1 , represented in terms of the zeros,
poles, and the transfer function gain K , where the static gain K s can be computed from
Ks G(1) .
n1
gij (q)
K (q qi )
i 0
m1
(q q )
i 0
i indicates the ith
output in a MIMO system.
76
ni.com
j indicates the jth
input in a MIMO system.
You then need to compute the polynomial roots of transfer function polynomials to convert a
transfer function model into a pole-zero model.
The Standard Deviation of Zeros and Poles
Refer to the si_Std Root to Polynomial.vi for more information about the implementing algorithm.
The si_Std Root to Polynomial.vi is a core VI used by the SI Poles-Zeros Plot.vi.
5.
Reference
si_Std Freq.vi
1.
Introduction
This section includes the algorithms and references used by the si_Std Freq.vi, which is the core
VI to estimate standard deviation of frequency response. The si_Std Freq.vi is located in the
labview\vi.lib\addons\System Identification\Model Presentation Subs.llb directory.
2.
You can use this VI to compute the confidence region of the Bode and Nyquist plots. The
frequency response function, G(e
coefficients
) , of a system can be computed from the estimated
of a linear polynomial model, such as an AR, ARX, ARMAX, OE, BJ or GL
model. Given that the standard deviation of
3.
is known, the purpose of this VI is to compute the
, and phase, arg G e , of the frequency
G e j
standard deviation of the magnitude,

response function,
G(e j ) .
APIs
Para Name
In/Out
Type
Description
freq response
In
1D CDB Array
Frequency response
In
1D DBL Array
Complex frequency response
transfer
In
Cluster
Alpha is the numerator of the transfer function.
containing
Beta1 * Beta2 is the denominator of the transfer
Alpha, Beta 1
function. Typically, Alpha is B(q), and Beta1 and
and Beta 2.
Beta2 are A(q) and F(q), respectively.
function
G(e j )
Alpha, Beta 1
77
and Beta 2 are

1D DBL Array.
covariance
In
2D DBL Array
Covariance matrix P of estimated model
matrix
coefficients N
standard
Out
deviation
Cluster
Standard deviation of frequency response
containing the
standard
deviation of the
magnitude and
phase response
4.
Algorithms
Since all the other polynomial models are converted to general-linear models when computing
frequency response, this section discusses only the computation formula for frequency response of
general-linear models.
Frequency Response of General Linear Model
The general-linear model (SISO) is defined as
A(q) y(t )
B(q)
C(q)
u(t )
e(t )
F (q)
D(q)
(1)
The frequency response between the input and output of the system is
G(e j )
B(e j )
A(e j )F (e j )
(2)
The general linear model (MISO) is defined as

N
A(q) y(t )
i 1
Bi (q)
C(q)
ui (t )
e(t )
Fi (q)
D(q)
(3)
The frequency response for i-th input and output is
Gi (e j )
Bi (e j )
A(e j )Fi (e j )
(4)
Gauss approximation formula [Page 292, 9-56, Ref 1]

Let
be an estimate of the coefficients of the polynomial model. The mean and covariance
matrix of
is
0 and P , respectively. If f (0 ) is the p -dimensional random variable,
asymptotically as
becomes sufficiently close to

^
0 , you can have
f () with good accuracy.
f ( ) f (0 ) f ' (0 )( 0 )
78
ni.com
where
f ' is the p d derivative of f

^
with respect to
^
. This means
(5)
The above expression is known as the Gauss approximation formula.

To solve the above function,
is the combination of the coefficients of A, B, and F.
magnitude or phase of frequency response function
G(e j ) .
f (0 )
is the
According to the Gauss
approximation formula, the covariance matrix of magnitude and phase of frequency response
function
G(e j )
are
P|G| Cov(| G(e j ) |) | G(e j ) | P | G(e j ) |T
(6)
and
PargG Cov(arg(G(e j ))) arg(G(e j )) P arg(G(e j ))T
Since
and
G(e j )
(7)
are known, you can apply Equation 6 and 7 to compute the standard
deviation of magnitude and phase of frequency response if you know the derivative of the
magnitude and phase of
G(e j )
arg(G(e j )) , you can derive
Derivative of
with respect to
. To derive
| G(e j ) |
and
G(e j ) as the following.
G(e j ) , | G(e j ) | and arg G(e j )
The derivative of
G(e j ) with respect to
with respect to
at frequency
is a vector defined as
G(e j1 ) [ G(e j1 ),
G(e j1 ),...,
G(e j1 ),
a1
a2
aNa
G(e j1 ),
G(e j1 ),...,
G(e j1 ),
b1
b2
bNb
(7)
G(e j1 ), G(e j1 ),...,

G(e j1 )]
f1
f2
f N f
where
G(e j1 ) j1 k
G(e j1 )
(e ) , k 1,2,..., Na ,
ak
A(e j1 )
G(e j1 ) j1 k
G(e j1 )
(e ) , k 1,2,..., Nb ,
bk
B(e j1 )
79
G(e j1 ) j1 k
j1
G(e )
(e ) , k 1,2,..., N f .
fk
F (e j1 )
The derivative of
G(e j )
is defined as

j1
G(e )
j2
G(e )
G(e j )
....
jN
G(e )

Since
(8)
G(e j ) G(e j ) G(e j ) , the derivative of | G(e j ) |
is
j
j
j
j
(G(e )). G(e ) (G(e )). G(e )
=
(G(e j )). G(e j )
j
G(e )
1
( G(e j ) )
2* G(e j )
where
G(e j )
derivative of
is the conjugate of
arg(G(e j ))
G(e j ) .
Because
(9)
arg G(e j ) ln G(e j ) , the
is
arg G(e j ) G(e j )./ G(e j )
By computing the derivative of
| G(e j ) |
and
arg G(e j ) with Equation 9 and 10 and then
applying Equation 6 and 7, you can get the covariance matrix of
| G(e j ) |
The diagonal of the covariance matrix is the variance of the vector
5.
(10)
and
| G(e j ) |
arg G(e j ) .
and arg G(e
).
Reference
si_Std Root to Polynomial.vi

1.
Introduction
This section includes the algorithms and references used by the si_Std Root to Polynomial.vi,
which is the core VI to calculate the confidence region of the roots of a polynomial. The si_Std
80
ni.com
Root to Polynomial.vi is located in the labview\vi.lib\addons\System Identification\Math Tools.llb

directory.
2.
You can use this VI to calculate the confidence region of the roots of a polynomial based on its
coefficients and the covariance of the coefficients.
3.
APIs
Para Name
In/Out
polynomial
Type
Description
In
1D DBL Array
The roots of a polynomial
In
2D DBL Array
A matrix containing the covariance of
coefficients
covariance
coefficients
std range
In
DBL
The factor that determines the confidence

region, which is std range * standard deviation
Roots
Out
1D CDB Array
circle plot data
Out
1D Array of
Each Cluster containing two 1D DBL Array,
Cluster
which represents the x-value and y-value of the

bound of a region
sta of root (special
Out
1D DBL Array
The variance of imaginary part of roots
format)
4.
Algorithms
Variance of the Transform [Page 292, 9-56, Ref 1]
Let the
d -dimension vector
be an estimate with mean
0 and its covariance matrix P . If

^
f (0 ) is the p -dimensional random variable, asymptotically as
0 , you can have
becomes sufficiently close to
f () with good accuracy.

^
f ( ) f (0 ) f ' (0 )( 0 )
where
f ' is the p d derivative of f

^
with respect to
^
. This means
^

This expression is known as the Gauss approximation formula.
Here, you view the roots of a polynomial as the function of its coefficients. So using the
covariance of coefficients and the derivative of roots with respect to coefficients, which is
calculated by the si_Polynomial Root Derivative.vi, you can get the covariance of roots.
81
Then by using the si_Variance of Real and Imaginary of Roots.vi, you can get the variance of the
real part and the imaginary part and their covariance for each root.
Set the variances you obtained as the input of the si_Confidence Region of Polynomial.vi, you can
get the confidence region of each roots.
5.
References
si_Polynomial Root Derivative.vi

1.
Introduction
This section includes the algorithms and references used by the si_Polynomial Root Derivative.vi,
which is the core VI to calculate the derivative of roots of a polynomial with respect to the
polynomial coefficients. The si_Polynomial Root Derivative.vi is located in the
labview\vi.lib\addons\System Identification\Model Presentation Subs.llb directory.
2.
Given a polynomial as represented by the following equation:
a0 xn ... an1x an a0 (x x1)(x x2 )...(x xn )
(1)
The purpose of this VI is to calculate the derivative of the roots with respective to the coefficients.
x1
x1
a a
n
0
xn xn
an
an
where
xi
xi
a
k
3.
are the roots and
ai
(2)
are the coefficients of the polynomial. However, this document uses
elsewhere to denote derivative matrix of Equation 2.
APIs
Para Name
In/Out
Type
Description
coefficients
In
1D DBL Array
The coefficients of a polynomial
dx/da
Out
2D CDB Array
The derivative of roots with respect to the

coefficients
4.
Algorithms
82
ni.com
First, consider that the given polynomial is a monic polynomial, where a0
1 for Equation 1,
with roots x1,..., xn .
xn b1xn1 ... bn1x bn x x1 ... x xn 0
(3)
From the Vieta Theorem, you can have
b1 (1) xi
i
b x x
2
i j
i j
bn (1)n x1...xn
(4)
or
bp (1) p
k1
...xk p , p 1,2,..., n
(5)
k1 ...k p
ki k j
You therefore can have
b1
1, q 1,2,..., n
xq
(6)
and
bp
(1) p xk1 ...xk p1
xq
k1 ...k p1
ki k j
ki q
p
(1) xk1 ...xk p1 xq xk1 ...xk p2
k1...k p1
.
k1 ...k p2
ki k j
ki k j
ki q
b
(1) p (1) p1 bp1 (1) p1 xq p1
xq
bp1 xq
(7)
bp1
xq
For a common polynomial
a0 xn ... an1x an a0 (x x1)(x x2 )...(x xn )
(9)
you can rewrite it as
xn b1xn1... bn1x bn (x x1)(x x2 )...(x xn )
(10)
where
bj
aj
, j 1,2,..., n .
a0
You therefore can have

83
xi
x
1 n
2 ak i
a0
a0 k 1 bk
xi 1 xi
, k 1, 2,..., n
ak a0 bk
(11)
(12)
si_Variance of Real and Imaginary of Roots.vi

1.
Introduction
This section includes the algorithms and references used by the si_Variance of Real and Imaginary
of Roots.vi, which is the core VI to calculate the variance of real part and imaginary part of
polynomial roots of a polynomial and their covariance. The si_Variance of Real and Imaginary of
Roots.vi is located in the labview\vi.lib\addons\System Identification\Model Presentation Subs.llb
directory.
2.
If the variance and covariance of a complex number and its conjugate are known, you can deduce
the variance of the real and imaginary parts and the covariance between the real and imaginary
parts. You can use this VI to calculate the variance of real and imaginary parts, the covariance of
real and imaginary parts if the roots of a polynomial are known, and the covariance matrix of the
roots. Note that the polynomial roots are required to be real or in conjugate pairs in this VI.
3.
APIs
Para Name
In/Out
Type
Description
polynomial roots
In
1D CDB Array
The roots of a polynomial, or real coefficients
Covariance
In
2D CDB Array
Covariance matrix of the polynomial roots
sta of root (special
Out
1D CDB Array
Reconstructing variance and covariance in an
format)
array, which has the same length to roots
Polynomial Roots
Out
1D CDB Array
Copy of polynomial roots
Re_Variance
Out
1D DBL Array
The variance of real part of the roots
Im_Variance
Out
1D DBL Array
The variance of imaginary part of the roots
Re_Im_Covariance
Out
1D DBL Array
The covariance of real part and imaginary part
dup
of the roots
4.
Algorithms
For a complex number,
can be written as
z , you can denote its real part as a
and its imaginary part as
b . Then, z
a bi . Thus, you can have
E( z) E(a) E(b)i ,
where
E( z) means the expectation value of z .
84
ni.com
Thus,
D( z) E( z E( z))2 E(a E(a))2 E(b E(b))2 2E(a E(a))(b E(b))i

D(a) D(b) 2Cov(a, b)i
where
D( z)
means the variance of
and
Cov(a, b)
means the covariance of
and
(1)
b.
Similarly, you can have
Cov( z, z) E( z E( z))( z E( z)) E(a E(a))2 E(b E(b))2

D(a) D(b)
where
(2)
z means the conjugate of z .
From Equations 1 and 2, you can get
1
Cov(a, b) (D( z))
2
(3)
1
D(a) (Cov( z, z) (D( z)))
2
If
1
D(b) (Cov( z, z) (D( z)))
2
z is a real number, you can treat z as a complex number with D(b) 0
(4)
(5)
and
Cov(a, b) 0 .
Then Equations 3, 4, and 5 become
D(a) D( z)
(6)
D(b) 0
(7)
Cov(a, b) 0
(8)
si_Confidence Region of Polynomial.vi

1.
Introduction
This section includes the algorithms and references used by the si_Confidence Region of
Polynomial.vi, which is the core VI to calculate the confidence regions of roots of a polynomial.
The si_ Confidence Region of Polynomial.vi is located in the labview\vi.lib\addons\System
Identification\Model Presentation Subs.llb directory.
2.
If you have knowledge about the variance and covariance of the real part and imaginary part of a
complex number, you can draw the confidence region of the complex number as an ellipse, or an
interval in the degenerate form, on the plane. Similarly, for a polynomial, if you know all the
above parameters of its roots, you can draw several ellipses on the plane to show the confidence
region of its roots. You can use this VI to calculate the bound of these ellipses.
3.
APIs
85
Para Name
polynomial
In/Out
Type
Description
In
1D CDB Array
var_Re
In
1D DBL Array
The variance of real part of roots
var_Im
In
1D DBL Array
The variance of imaginary part of roots
covar_ReIm
In
1D DBL Array
The covariance of real part and imaginary part of
roots
roots
num of points
In
I32
The number of points in each circle
std range
In
BDL
The factor determines the confidence region,

which is std range * standard deviation.
circle plot data
Out
1D Array of
Each Cluster contains 2 1D DBL Array, which
Cluster
represent the x-value and y-value of the bound of a

region.
4.
Algorithms
For a complex number,
X , you can denote its real part and imaginary part as x1 and x2 .
Suppose that their average values are
1, 2 , their variances are 1, 2 , and their covariance is
1 2 , where 0 1 .
If one of
and
is zero, then one part of
interval on another direction with mid-point
X is a constant, and the confidence region is an
(1, 2 )
and radius std range* i , which does not
equal 0.
Otherwise, you can consider the distribution of
P( x1, x2 )
x1
and
x2
as bivariate normal distribution
z
exp
2
2 1 2
21 2 1
1
(1)
where
2
2
x1 1 2 x1 1 x2 2 x2 2
12
1 2
22
(2)
You can move the average point to zero and rotate the axis as shown in the following equation by
coordinate converting:
x1 1 r1 cos r2 sin
x2 2 r1 sin r2 cos
Then,
86
ni.com
cos2 2 cos sin sin 2 2

2 r1
2
1 2
2
1
cos sin cos2 sin 2 cos sin

r1r2
2
12
1 2
22
2
2
sin 2 cos sin cos 2
2
r2
2
1 2
2
1
Let
(3)
r2 0 , so you can have the conditional distribution
cos2 2 cos sin sin 2

2
2
1
1 2
2 2
P(r1 | r2 0) const exp

r1
2
2 1
(4)
The constant before the exponential function is not necessary to know, but you must know it is
also a normal distribution with
cos2
12
1
2 cos sin
1 2
1 2
sin2
22
(5)
is the variance.
So, in this direction, (cos ,sin ) , the confidence region is [-std range* , std range* ]. Thus,
where
you have a point on the ellipse as (std range* * cos + 1 , std range* * sin + 2 ). You can
get the ellipse by connecting all these points.
For a polynomial, you can handle all its roots following the above method.
SI Average Theta and Covariance.vi

1.
Introduction
This section includes the algorithms and references used by the SI Average Theta and
Covariance.vi. The SI Average Theta and Covariance.vi is the core VI used to merge general linear
polynomial models. The SI Average Theta and Covariance.vi is located in the
labview\vi.lib\addons\System Identification\Model Merge.llb directory.
2.
The purpose of this function is to merge general-linear polynomial models estimated from
different experimental data sets. The models to be merged must be of the same model type and
must have the same model orders.
87
3.
APIs
Para Name
In/Out
Type
Description
theta 1
In
2D DBL Array
The parameters of model 1
covariance 1
In
3D DBL Array
The covariance matrix of model 1
theta 2
In
2D DBL Array
The parameters of model 2
covariance 2
In
3D DBL Array
The covariance matrix of model 2
theta
Out
2D DBL Array
The parameters of the merged model
covariance
Out
3D DBL Array
The covariance matrix of the merged

model
4.
Algorithms
This VI uses
i to denote the parameters estimated from dataset i and C i
to denote the
corresponding covariance matrix. The optimal way to combine these estimates is demonstrated by
the following equation: [Ref 1]
2
C C i
i1
1 1
5.
i
1
CC i
(1)
i 1
References
[1] Ljung, L. 1999. System Identification Theory for the User. 2nd ed. Prentice Hall, Page 465,
Equation 14.15.
SI Generate Pseudo-Random Binary Sequence.vi

1.
Introduction
This section includes the algorithms and references used by the SI Generate Pseudo-Random
Binary Sequence.vi. The SI Generate Pseudo-Random Binary Sequence.vi is the core VI used to
generate pseudo-random binary sequences (PRBS). The SI Generate Pseudo-Random Binary
Sequence.vi is located in the labview\vi.lib\addons\System Identification\Utility.llb directory.
2.
The purpose of this function is to generate a PRBS to be used as the stimulus signal of a system.
3.
APIs
Para Name
In/Out
Type
Description
samples
In
I64
The length of the sequence
polynomial order
In
I32
The order of the primitive polynomial
seed
In
I64
The seed of a PRBS
88
ni.com
4.
upsampling factor
In
I32
The upsampling factor n
PRBS
Out
1D U8 Array
The PRBS this VI generates
Algorithms
High frequency fluctuations can be harmful to devices in the real-world industrial environment.
Therefore, a stimulus signal whose power spectrum is similar to a lowpass filter is often desirable.
You can obtain a lowpass PRBS by upsampling the original PRBS, also known as the original
maximum length sequence, whose spectrum is uniform over the entire frequency range. [Ref 1]
n
where n is the upsampling factor,
Tclk
Tclk
Ts
(1)
is the clock period of the original maximum length
sequence, and Ts is the sampling period.
5.
Reference
[1] Yucai Zhu, Multivariable System Identification for Process Control, Elsevier Science, 2001,
Page 43.
Gauss-Newton Minimization Method

1.
Introduction
This section summarizes the algorithms of the Gauss-Newton minimization method. You can
use the Gauss-Newton minimization method to refine the estimation of continuous transfer
function models, user-defined models, partially known models, and polynomial models such
as the ARMAX, output-error (OE), Box-Jenkins (BJ), and general-linear models.
2.
Problem to Solve
The purpose of model estimation is to identify the model coefficients by minimizing the mean
square error,
VN , which is defined by the following equation:

VN ( )
1 N 1 2
(t, )
N t 1 2
(1)
In this equation, all the coefficients to be estimated are combined together as a vector.
(t , )
is the error
y (t, ) y(t ) ,
simulated output of the system,
indicating the difference between the predicted or
yt, , and the measured output, yt .
You can use some methods, such as the multi-stage method for polynomial models, to get a
coarse estimation of
. You then can use the following iteration to refine :

89
(i1) (i ) f (i )
where
is the step size and
(2)
is the search direction. The purpose of iteration is to
f (i )
minimize the mean square error VN
. The problem to solve lies in how to select and
compute the search direction.

3.
Algorithms [page 326, Ref 1]

The Gauss-Newton minimization defines the search direction
f ( ) as
f ( ) V ''( ) V '( )
1
(3)
where
1 N
(t, ) (t, )
N t 1
(4)
1 N
(t, ) T (t, )
N t 1
(5)
V '( )
and
V ''( )
t,
dyt,
dt
(6)
T denotes transposing. By inserting Equations 4 and 5 into 3, you obtain the following
equation:
N
f ( )
(t, ) (t, )
t 1
N
(t, )
(t, )
t 1
(1, )
(2, )
(1, ) (2, ) ... ( N , ) ...
( N , )
T (1, )
T
(2, )
(1,
(2,
)
...
(
N
,
...
T
( N , )
T (1, )
T
(2, )
...
T
( N , )
You can evaluate
f ( )
(1, )
(2, )
...
( N , )
(7)
by solving the following linear equation:
90
ni.com
T (1, )
(1, )
T
(2, ) f ( ) (2, )
...
...
T
( N , )
( N , )
(8)
The following equations define (t , ) .

Gradient of ARMAX Model [Page 330, Equation 10.52, Ref 1]
(t, )
1
[ y(t 1)... y(t Na ) u(t 0)...u(t Nb 1)
C(q)
(t 1)... (t Nc )]T
Gradient of OE Model [Page 330, Equation 10.54b,e, Ref 1]
1 k
Bq
q u(t ),
q k ut , t 1,...,N
F qF q
F q
t,
Gradient of BJ Model [Page 330, Equation 10.54b,c,d,e, Ref 1]
D(q)
D(q) k
qk u(t ),
q w(t ),
C(q) F (q)
C(q) F (q)
1
1 k
qk (t ),
q v(t )], t 1,..., N
C(q))
C(q))
(t, ) [
Gradient of General-Linear Model [Page 330, Equation 10.54a,b,c,d,e, Ref 1]
(t, ) [
D(q) k
D(q)
D(q) k
q y(t ),
qk u(t ),
q w(t ),
C(q)
C(q) F (q)
C(q) F (q)
1
1 k
qk (t ),
q v(t )], t 1,..., N
C(q))
C(q))
Gradient of Continuous Transfer Function Models, User-Defined Models, and Partially

Known Models
The LabVIEW System Identification Toolkit does not have explicit formulas to calculate the
gradients for continuous transfer function, user-defined, or partially known models. Instead,
this toolkit uses the following numerical differentiation equation to calculate the gradients.
yt, d yt, d
2 d
d *DIFF _ 2
t,
where
4.
DIFF _ 2 = 3.2057501263e6
Robustness Consideration [Pages 481-483, Ref 1]

Since the optimization criterion is to minimize the norm of the prediction error as shown by
Equation 1, accurate estimation of the norm of the prediction error can improve the robustness
of the minimization. The prediction error is clipped using the following equation: [Page 483,
Equation(15.9), Ref 1]
91
(n) | (n) |
(n) (n)
(n)
where p is a scalar in the range

selects
1.5 .
1 1.8
. The System Identification Toolkit
is the estimated standard deviation of the prediction error by using the
MAD method, as shown by the following equation:
MAD* 1.483263287118
where MAD = the median of { | (t ) ~ | } with ~ as the median of (t ) [Page 483, Ref 1].
The constant 1.483263287118 is obtained with
5.
Reference
Instrumental Variable Method

1.
Introduction
This document includes the algorithm summary on the instrumental variable method to identify
an ARX model.
2.
Problem to Solve
An ARX model can be defined as

If the noise e(n) is uncorrelated to the regression variables, y(n) and u(n), the coefficients of the
model A(q) and B(q) can be estimated with least squares regression method. However, if the
noise is correlated to y(n) and u(n), the least squares regression method does not work well. The
instrumental variable method described in this document can solve this problem.
3.
Algorithms
The instrumental variable method suggests the solution for the ARX model as [Eq. 7.118, P. 224,
Ref 1]:
NLS [
where
N
1 N
T
1 1
(
t
)
(
t
)
]
(t) y(t)
N t 1
N t 1
(1)
contains the instruments variables.
Choice of Instruments Variables
[P.225, Ref 1, 3.2, Ref 2]
92
ni.com
For an open-loop system, the input sequence of the system, or its filtered version, is often a good
choice of instrumental variables. One of the choices of instrumental variables is
[x(t 1) x(t 2).... x(t Na ) u(t) u(t 1)...u(t Nb 1)]T

where
x(t )
(2)
is generated from the input through a linear system

c N (q) x(t ) M (q)u(t )
(3)
where
N (q) 1 n1q 1 n2 q 2 .... nnN q nN
(4)
M (q) 1 m1q 1 m2 q 2 .... mmN q mN
(5)
In practice, you can use the least squares method to estimate the coefficients of an ARX model.
You then can use the estimated A(q) and B(q) as N(q) and M(q), respectively, to have an
instrumental variable estimation of the ARX model with Equation 1.
4.
Reference
[2] National Instruments xMath User Manual.
SI Estimate Discrete Transfer Function Model from

FRF (SISO arbitrary).vi
1.
Introduction
This section includes the algorithms and references used by the SI Estimate Discrete Transfer
Function Model from FRF (SISO arbitrary).vi. The SI Estimate Discrete Transfer Function Model
from FRF (SISO arbitrary).vi is the core VI used to estimate discrete transfer functions from FRF
data. The SI Estimate Discrete Transfer Function Model from FRF (SISO arbitrary).vi is located in
the labview\vi.lib\addons\System Identification\Frequency Domain Estimation.llb directory.
2.
The discrete transfer function model can be represented as
y( k ) H( z )u( k ) e( k ) ,
(1)
where u(k) and y(k) are the stimulus and response signals of a system, e(k) is the additive noise,
and H(z) is represented by the following equation.
H( z )
b0 b1z... bm z
a0 a1z... an z
, mn
(2)
where m is the numerator order and n is the denominator order.

The purpose of this VI is to estimate the transfer function H(z) by using the frequency response
function (FRF) of a system.
93
3.
APIs
Para Name
In/Out
Type
Description
FRF magnitude
In
Cluster
The magnitude of the FRF of the system
FRF phase
In
Cluster
The phase of the FRF of the system
FRF format
In
Cluster
orders of
In
Cluster
transfer function
denominator order
model
FRF weight
In
1D CDB Array
The frequency-domain weight of the

estimation error
sampling rate
In
DBL
system model
Out
Object containing
out
information about
this model
coefficients of
Out
Cluster
Numerator is [b0, b1, , bm],
transfer function
Denominator is [a0, a1, , an]
model
4.
Algorithms
This VI uses the polynomial Matrix Fraction Description (MFD) to represent the transfer function
model. The polynomial MFD also estimates MIMO models. This estimation method is a
multivariable least squares method in the frequency domain.
For a set of observed FRF data G( ) , evaluated at N frequency points
j
p r
i
for j 1,...,N
G( e ) G( j ) C
(3)
The aim of the identification procedure is to find a linear time-invariant (LTI) model P( ) that
approximates the above FRF data. The model P( ) is parameterized by either a left or right
polynomial MFD, where collects all the real-valued parameters of the polynomial MFD.
The model P( ) approximates the data G( ) by minimizing the following equation.
j
J( )
trE( j , )E * ( j , ) E( ) 2F
N
(4)
j 1
where E( , ) G( ) P( , ) for j 1,...,N .

j
The goal of this algorithm is to find a real-valued in the following equation.
: argminJ( )
(5)
Focusing on the left-hand side of the polynomial MFD, the model P( ) can be parameterized as
94
ni.com
P( z , ) A( z , ) B( z , )
(6)
where A and B denote parameterized polynomial matrices in terms of z-1. Note that
P( j , ) P( z , )
(7)
i jT
z e
For a model with r inputs and p outputs, the following equations represent A and B, respectively:
1
A( z , ) I p p
where A
p p
pr
(8)
k 1
d m 1
k d
Ak z
B( z , )
where B
Bk z
(9)
and d denotes the number of leading zero matrix coefficients. Because each
input-output pair defines a SISO LTI system, Equations 8 and 9 are scalar polynomials.
When the system is defined by scalar polynomials, you can determine dij by the number of time
delays from the j-th input to the i-th output. These pair-wise parameters, dij, mij, and nij are
structural parameters and are set to d:=min{dij}, m:=max{mij}, and n:=max{nij} for Equations 8
and 9 in the general MIMO case.
Then, the error E( , ) is defined by the following:
j
1
1
1
1
E( j , ) A( z , ) A( z , )G( j ) B( z , ) i T
z e j
(9)
From here, a multivariable case SK-iteration can be invoked to actually minimize the error. If
necessary, a subsequent GN-optimization can be utilized by using the SK-iteration result as the
initial estimate.
Because this VI applies to SISO systems only, the above algorithm is used in a degenerated case,
where r=p=1.
5.
References
[1] R. A. Callafon, D. de Roover, and P. M. J. Van den Hof, Multivariable least squares frequency
domain identification using polynomial matrix fraction descriptions, in Proc. 35th Int. Conf.
Decision and Control, Kobe, Japan, Dec. 1996, pp. 2030-2035.
95
SI Estimate Continuous Transfer Function Model from

FRF (SISO arbitrary).vi
1. Introduction
This section includes the algorithms and references used by SI Estimate Continuous Transfer
Function Model from FRF (SISO arbitrary).vi. The SI Estimate Continuous Transfer Function
Model from FRF (SISO arbitrary).vi is the core VI used to estimate continuous transfer function
model from FRF data. The SI Estimate Continuous Transfer Function Model from FRF (SISO
arbitrary).vi is located in the labview\vi.lib\addons\System Identification\Frequency Domain
Estimation.llb directory.
2. Function Description
The continuous transfer function model can be represented as
y(k ) H (s)u(k ) e(k )
(1)
where u(k) and y(k) are the input and output of a system, e(k) is the additive noise, and H(s) is
H ( s)
b0 b1s... b j s ... bm s
a0 a1s... ai s ... an s
,m<n
(2)
where m is the numerator order and n is the denominator order. The purpose of this VI is to
estimate the transfer function H(s) by using the FRF of a system.
3. APIs
Para Name
In/Out
Type
Description
FRF magnitude
In
Cluster
The magnitude of the FRF of a system
FRF phase
In
Cluster
The phase of the FRF of a system
FRF format
In
Cluster
orders of
In
Cluster
transfer function
denominator order
model
FRF weight
In
1D CDB Array
The frequency-domain weight of the

estimation error
system model
Out
out
Object containing
model
coefficients of
Out
Cluster
Numerator is [b0, b1, , bm],
transfer function
Denominator is [a0, a1, , an]
model
4. Algorithms
96
ni.com
The algorithm used in this VI is a least squares method. To have better numerical properties, the
basis polynomials are chosen to be the Chebyshev polynomials[1] instead of the natural basis
polynomials {1, s, s2, , sn}.
Chebyshev polynomials are defined in many ways. One way by recursion is as follows:
T0(s) = 1
T1(s) = s
Tn+1(s) = 2sTn(x) Tn-1(s)
where the resulting basis polynomials are {T0, T1, T2, , Tn}.
Now you can define H(s) as:
H( s )
b0 b1 T1( s ) ... b j T j ( s ) ... bmTm ( s )

c
(3)
a0 a1 T1( s ) ... ai Ti ( s ) ... anTn ( s )
Suppose you obtain a set of observed FRF data G( ) , evaluated at N frequency points .
j
G( ) G( ) C , for j 1,...N
j
j
By evaluating H(s) at the same set of N frequency points , you reach a standard setup for a least
j
squares problem:
N
H( , j ) G( j )
argmin
j 1
where H( , ) H( s )
j
siw j
and a0
a1
... an b0 b1
c
... bm .
After obtaining , the coefficients under the Chebyshev basis, you can insert them back into
Equation 3 so as to obtain the coefficients {a0, a1, , an, b0, b1, , bm}, which are under the natural
basis.
5. References
[1] Vetterling, William T., William H. Press, Saul A. Teukolsky, and Brian P. Flannery. 1992.
Numerical Recipes in C: The Art of Scientific Computing. Cambridge University Press
[2] Xmath System Identification Module, Part 1
SI Estimate State-Space Model from FRF.vi

1.
Introduction
This section includes the algorithms and references used by the SI Estimate State-Space Model
from FRF.vi. The SI Estimate Discrete State-Space Model from FRF (MIMO arbitrary).vi is the
core VI used to estimate a discrete state-space model from FRF data. The SI Estimate State-Space
97
Model from FRF.vi is located in the labview\vi.lib\addons\System Identification\Frequency

Domain Estimation.llb directory.
2.
The discrete state space models in this VI are defined as
x( k 1) Ax( k ) Bu( k )
y( k ) Cx( k ) Du( k ) n( k )
m
(1)
where u( k ) , y( k ) , x( k ) are the input, output, and state, respectively, of the

l
system. n( k ) is the noise term which is assumed to be independent of the input sequence
u(k). By using the Fourier transform on both sides of both equations in Equation 1, you obtain the
following equations.
i
e X ( ) AX ( ) BU( )
Y ( ) CX( ) DU( ) N( )
where Y(), U(), N(), and X() are the transformed output, input, noise, and state,
respectively. By eliminating the state, you obtain the following equation.
Y ( ) G( e
i
where G( e
) G( z )
z e
D C( zI A ) B
z e
)U( ) N( )
and is the frequency response function
(FRF) of the system.

The goal of this VI is to find an estimate for the system matrices A, B, C, and D, and the initial
states x0, given samples of the FRF G( e
3.
) of the system.
APIs
Para Name
In/Out
Type
Description
FRF magnitude
In
Cluster
The magnitude of the FRF of the system
FRF phase
In
Cluster
The phase of the FRF of the system
FRF format
In
Cluster
number of states
In
I32
Denoted by n, also known as the order of

the system
noise subspace
In
DBL
Determines the augmented order to

incorporate noise at a given number of
states
structure selector
In
Cluster
Information about the D matrix and initial

state
sampling rate
In
DBL
system model
Out
Object containing
98
ni.com
out
information about
this model
coefficients of
Out
Cluster
Estimated matrices A, B, C, D
Out
1D DBL Array
Estimated initial state vector
state-space
model
initial states
4.
Algorithms
The algorithm used in this VI is a frequency-domain subspace identification method, which is very
similar to its time-domain counterpart.
y( k )
y( k 1 )
and similarly with u ( k ) and n ( k )
Define yd ( k )
d
d
y( k d 1 )
By recursively using Equation 1, you obtain:
yd d x( k ) d ud ( k ) nd ( k )
(2)
C
CA
and is the extended system observability matrix and
where d

d 1
CA
0 0 0
D
CB
D 0 0
d CAB CB D 0

d 2
CA B CB
0
0
0 is the lower triangular Toeplitz matrix.
0
D
By using the Fourier transform on both sides of Equation 2, we obtain
Wd ( ) Y( ) d X ( ) dWd ( ) U( ) Wd ( ) N( )
where denotes the Kronecker product and Wd ( ) 1 e
2i
... e
(3)
i( d 1 ) T
Assume you have samples of the Fourier transform of the input and output signals at M
frequencies k, where k=1, 2,, M. By collecting these samples properly, you have
Yd ,M (Wq ( 1 ) Y( 1 ) ... Wq ( M ) Y( M ) )
Ud ,M (Wq ( 1 ) U( 1 ) ... Wq ( M ) U( M ) )
Nd ,M (Wq ( 1 ) N( 1 ) ... Wq ( M ) N( M ) )
X M ( X ( 1 ) ... X ( M ) )
By using Equation 3, you obtain
99
Yd ,M d X M dU d ,M Nd ,M
(4)
From Equation 4, you want to remove the term U

by some proper orthogonal projection.
d d ,M
Next, you can estimate d, which will enable you to obtain estimates of matrices A and C. You
then can estimate matrices B and D and the initial state x0 by using the least squares method.
This VI uses the following five steps to implement the algorithm described above by employing
QR factorization and singular value decomposition (SVD) for efficiency and reliability.
1) Form the matrices Y and U in Equation 4. You can suppress the subscript variables for
simplicity. Notice that because you begin with the FRF, you can assign U to be the unit or the
Fourier transform of the impulse signal and directly relate FRF and Y.
2) Calculate the QR factorization
H
U R11 0 Q1
R
Y 21 R22 Q H
2
3) Calculate the SVD of R22
ReR
22
Im R22 U S U O S
0
T
0 VS
O V T
O
4) Determine A and C:
( J U ) J U
A
1 s
2 s
C J 3U s
where J ( I
1
( d 1)l
0( d 1)l l ) , J 2 ( 0( d 1)l l
I( d 1)l ) , and J3 ( Il 0l ( d 1)l ) and are
the selection matrices.

5) Solve for B, D, and x0 from the following least squares problem
M
,D
, x argmin
B
0
5.
k 1
Y ( k ) ( D C ( e
ik
)1 B )U ( ) C ( e
IA
k
ik
)1 x
IA
0
References
[1] McKelvey, T. 1995. Identification of State-Space Models from Time and Frequency Data.
Ph.D. thesis, Department of Electrical Engineering, Division of Automatic Control, Linkoping
University, Sweden.
100
ni.com

Si Reference

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Si Reference

Încărcat de

Drepturi de autor:

Formate disponibile

LabVIEW

System Identification Toolkit Algorithm References

WARNING REGARDING USE OF NATIONAL INSTRUMENTS PRODUCTS

National Instruments Corporation

System Identification Toolkit

System Identification Palettes, VIs, and SubVIs

SI Estimate Discrete Transfer Function Model

SI Estimate State-Space Model

si_AR (FB, LS, YW and PC).vi

SI Estimate ARX Model

System Identification Toolkit

si_Convert to Transfer Function Model.vi

SI Convert to Models of CDT

si_Convert to State Space.Model.vi

SI Model Residual Analysis

si_Piecewise Linear Fitting.vi

The piecewise points include 0, x1, , xL,

The signal to be fitted

The piecewise points of the linear fitting

The best fit piecewise linear signal

National Instruments Corporation

System Identification Toolkit

The piecewise slope

The piecewise intercept

respectively, the best fit value of

solving the following linear equations:

Y [ y0, y1 yL , y N 1 ]T , S [s0, s1 , s N 1 ]T , and

and mi represents the number of signals between

System Identification Toolkit

SI Detect Feedback (Array).vi

The input signal of a system

The output signal of a system

The number of points for which t<0 and

Determines the upper and lower limits of

the impulse response

National Instruments Corporation

System Identification Toolkit

to N*std(-j), where N is determined by the confidence level, h(j) is considered to be zero. If

si_AR (FB, LS, YW and PC).vi

A(q) y(n) e(n)

white noise, and y(n) is the signal. q is the backward shift

operator, which means:

The signal to be fitted with Equation 1

Number of auto-regression coefficients, Na

U16 Text Ring

Select the method to estimate the AR model.

[1, a1, a2 ,....,aNa ]

[1, a1, a2 ,....,aNa ]

System Identification Toolkit

where N is the signal length.

National Instruments Corporation

System Identification Toolkit

si_AR Burg Method.vi

A(q) y(n) e(n)

operator, which means:

System Identification Toolkit

The input signals of a system

The output signal of a system

A order, B order, and delay are

order, B order, and

[N1b, N2b,...., NNb] , and [k1, k2,...., kN ] ,

I32. B order and

The standard deviations of coefficients A and B

x f (n) akf (k ) x(n k )