ISA

TRANSACTIONS®
ISA Transactions 41 共2002兲 511–520

Design PID controllers for desired time-domain

or frequency-domain response
Weidong Zhang,* Yugeng Xi, Genke Yang, Xiaoming Xu
Department of Automation, Shanghai Jiaotong University, Shanghai 200030, PRC
共Received 10 July 2001; accepted 8 February 2002兲

Abstract
Practical requirements on the design of control systems, especially process control systems, are usually specified in
terms of time-domain response, such as overshoot and rise time, or frequency-domain response, such as resonance peak
and stability margin. Although numerous methods have been developed for the design of the proportional-integral-
derivative 共PID兲 controller, little work has been done in relation to the quantitative time-domain and frequency-domain
responses. In this paper, we study the following problem: Given a nominal stable process with time delay, we design
a suboptimal PID controller to achieve the required time-domain response or frequency-domain response for the
nominal system or the uncertain system. An H ⬁ PID controller is developed based on optimal control theory and the
parameters are derived analytically. Its properties are investigated and compared with that of two developed suboptimal
controllers: an H 2 PID controller and a Maclaurin PID controller. It is shown that all three controllers can provide the
quantitative time-domain and frequency-domain responses. © 2002 ISA—The Instrumentation, Systems, and Auto-
mation Society.

Keywords: Linear system; Time delay; PID controller; Optimal control; Time-domain response

1. Introduction the focus of the study is on the optimal or subop-

Most industrial processes are controlled by
proportional-integral-derivative 共PID兲 controllers.
The popularity of PID controllers can be attributed
partly to their robust performance in a wide range
of operating conditions and partly to their func-
tional simplicity, which allows process engineers
to operate them in a simple and straightforward
manner. To implement such a controller, three pa-
rameters must be determined for the given pro-
cesses, and many papers have been devoted to this
problem. For example, see 关1– 6兴. Early study on
the design and tuning of a PID controller is mainly
based on classical control theory. In recent years,
*Corresponding author. Tel: ⫹86.21.62933329;
fax: ⫹86.21.62826946. E-mail address:
wdzhang@mail.sjtu.edu.cn
timal design of PID controllers. 关7兴 and 关8兴 pre-
sented a PID controller by Padé approximation in
the framework of internal model control 共IMC兲. In
fact, the controller is an H 2 suboptimal controller.
关9兴 provided a PID controller by approximating an
IMC controller with a Maclaurin series. 关10兴 pro-
vided a suboptimal PID controller for integrator/
time delay processes. The superiority of these
methods is that the controller is derived analyti-
cally. The control system can be designed and
tuned by formulas and thus the design and tradeoff
are significantly simplified.
Although numerous methods have been devel-
oped, little work has been done in relation to the
quantitative time-domain and frequency-domain
responses. As we know, practical requirements on
the design of control systems, especially on pro-
cess control systems, are usually specified in terms

0019-0578/2002/$ - see front matter © 2002 ISA—The Instrumentation, Systems, and Automation Society.
512 Weidong Zhang et al. / ISA Transactions 41 (2002) 511–520

Fig. 1. Unity feedback control loop.

Fig. 2. IMC structure.

of time-domain response, such as overshoot and
rise time, or frequency-domain response, such as
resonance peak and stability margin. It can be sim- C共 s 兲
ply formulated as follows: Given a nominal plant Q共 s 兲⫽
1⫹G 共 s 兲 C 共 s 兲
and desired time-domain response or frequency-
domain response, we design a PID controller to and we have
meet these indexes. In this paper, an H ⬁ PID con-
troller is first presented based on optimal control Q共 s 兲
C共 s 兲⫽ . 共2兲
theory. The properties of the controller are then 1⫺G 共 s 兲 Q 共 s 兲
investigated and compared with an H 2 PID con-
troller and a Maclaurin PID controller. It is shown In the nominal case, the unity feedback loop can
that these three controllers can provide quantita- be equivalent to the well-known IMC structure de-
tive time-domain response and frequency-domain picted in Fig. 2. The sensitivity transfer function
response. The work is of significance in that it can of the closed-loop system can be written as
establish a relationship between the classical de- 1
sign requirement and that of optimal design meth- S共 s 兲⫽ ⫽1⫺G 共 s 兲 Q 共 s 兲 共3兲
1⫹G 共 s 兲 C 共 s 兲
ods, and provide insight into control system de-
sign. and the complementary sensitivity transfer func-
The paper is organized as follows. In Section 2 tion as
the suboptimal H ⬁ PID controller is derived ana-
lytically. The H 2 PID controller and Maclaurin G共 s 兲C共 s 兲
T共 s 兲⫽ ⫽G 共 s 兲 Q 共 s 兲 . 共4兲
PID controller are briefly introduced. The perfor- 1⫹G 共 s 兲 C 共 s 兲
mance degree is defined and the quantitative time-
T ( s ) is just the closed-loop transfer function.
domain response and frequency-domain response
It is well known that any model obtained by
are discussed in Section 3. Finally, the paper is
using experimental procedures involves uncer-
concluded in Section 4 with a brief discussion of
tainty in its parameters. For the low-order models
the results.
used in many practical applications, the uncer-
tainty is expressed in a parametric fashion. The
2. Controller design parameters of the simplified model used to ap-
proximate a usually high-order dynamics are
Consider the unity feedback control system known to lie in an interval. The mean value of
each parameter is chosen to represent its nominal
shown in Fig. 1, where G̃ ( s ) denotes the plant and
value, often used to obtain nominal plant and con-
C ( s ) denotes the controller. In the context of pro-
troller parameters 关8,9兴. For example, the plant is
cess control, the process G̃ ( s ) is usually described
by the following model 共or the nominal plant兲 ˜
K̃e ⫺ ␪ s
G(s): G̃ 共 s 兲 ⫽ 共5兲
˜␶ s⫹1
Ke ⫺ ␪ s
G共 s 兲⫽ , 共1兲 with
␶ s⫹1
K̃苸 关 K ⫺ ,K ⫹ 兴 , ˜␪ 苸 关 ␪ ⫺ , ␪ ⫹ 兴 , ˜␶ 苸 关 ␶ ⫺ , ␶ ⫹ 兴 .
where K is the gain, ␶ is the time constant, and ␪ is
the time delay. We define the transfer function The midrange plant is chosen as the model:
 Weidong Zhang et al. / ISA Transactions 41 (2002) 511–520 513

Ke ⫺ ␪ s In other words, we must guarantee that S ( s )

G共 s 兲⫽ , 共6兲 ⫽ ( 1⫺G ( s ) Q ( s )) has a zero at s⫽0 to cancel the
␶ s⫹1
pole of W ( s ) . With the constraint, the unique op-
where timal Q im ( s ) is obtained as follows:

K ⫺ ⫹K ⫹ ␪ ⫺⫹ ␪ ⫹ ␶ ⫺⫹ ␶ ⫹ 共 ␶ s⫹1 兲共 1⫹ ␪ s/2兲
K⫽ , ␪⫽ , ␶⫽ . Q im 共 s 兲 ⫽ .
2 2 2 K

A central objective in automatic control is that a Obviously Q im ( s ) is improper. Now we use the
physical quantity is made to behave in a pre- following low-pass filter to roll Q im ( s ) off at high
scribed way by using the error between the system frequency:
output and the setpoint input. This gives rise to the 1
optimal control. Assume that the optimal perfor- J共 s 兲⫽ , ␭⬎0.
mance index is H ⬁ optimal, i.e., min储W(s)S(s)储⬁ , 共 ␭s⫹1 兲 2
where W ( s ) is a weighting function. W ( s ) should Then
be selected such that the 2-norm boundary of the
system input is normalized by unity. That is, 共 ␶ s⫹1 兲共 1⫹ ␪ s/2兲
W ( s ) ⫽1/s for a unit step setpoint. The perfor- Q 共 s 兲 ⫽Q im 共 s 兲 J 共 s 兲 ⫽ .
K 共 ␭s⫹1 兲 2
mance index implies that the controller is designed 共11兲
to minimize the worst error resulting from system
inputs. With Padé approximation, we have As ␭ tends to be zero, the controller tends to be
optimal. The corresponding controller of the unity
1⫺ ␪ s/2 feedback loop is
G 共 s 兲 ⫽K , 共7兲
共 ␶ s⫹1 兲共 1⫹ ␪ s/2兲 1 共 ␶ s⫹1 兲共 1⫹ ␪ s/2兲
C共 s 兲⫽ . 共12兲
which will be regarded as the nominal plant uti- Ks ␭ 2 s⫹2␭⫹ ␪ /2
lized to derive the H ⬁ PID controller. The error
introduced by the approximation is included in un- The suboptimal H ⬁ PID controller is derived ana-
certainty, which will be discussed later. lytically. Comparing the above controller with the
Instead of a numerical method, an analytical de- following practical PID controller,
sign procedure is developed for the given plant. It
is seen that W ( s ) S ( s ) is analytical in the open
right half plane. According to the well-known
冉
C 共 s 兲 ⫽K C 1⫹
1
T Is
⫹T D s 冊 1
T F s⫹1
. 共13兲

maximum modulus theorem, a fundamental fact The parameters of the new PID controller are
concerning complex functions, 兩 W ( s ) S ( s ) 兩 does
not attain its maximum value at an interior point ␭2 ␪ ␪␶
T F⫽ , T I⫽ ⫹ ␶ , T D⫽ ,
of the open right half plane. On the other hand, the 2␭⫹ ␪ /2 2 2T I
G ( s ) has a zero at s⫽2/␪ in the open right half
plane. Thus, for all Q ( s ) ’s, TI
K C⫽ .
K 共 2␭⫹ ␪ /2兲
储 W 共 s 兲共 1⫺G 共 s 兲 Q 共 s 兲兲 储 ⬁ ⭓ 兩 W 共 2/␪ 兲 兩 . 共8兲
If the practical PID controller is in the form of

冉 冊
Consequently we have
1 T D s⫹1
min储 W 共 s 兲 S 共 s 兲 储 ⬁ ⫽min储 W 共 s 兲共 1⫺G 共 s 兲 Q 共 s 兲兲 储 ⬁ C 共 s 兲 ⫽K C 1⫹ , 共14兲
T I s T F s⫹1
⫽ ␪ /2. 共9兲 the parameters of the PID controller are
However, W ( s ) has a pole on the imaginary axis. ␭2 ␪
To obtain a finite infinity norm, a constraint will T F⫽ , T I⫽ ␶ , T D⫽ ,
2␭⫹ ␪ /2 2
be imposed on the design procedure:
lim 共 1⫺G 共 s 兲 Q 共 s 兲兲 ⫽0. 共10兲 TI
K C⫽ .
s→0 K 共 2␭⫹ ␪ /2兲
514 Weidong Zhang et al. / ISA Transactions 41 (2002) 511–520

If the practical PID controller is in the form of ␭␪ ␪ ␪␶

T F⫽ T I⫽ ⫹ ␶ , T D⫽
冉 冊
, ,
1 T Ds 2 共 ␭⫹ ␪ 兲 2 2T I
C 共 s 兲 ⫽K C 1⫹ ⫹ , 共15兲
T I s T F s⫹1
TI
K C⫽ .
the parameters of the PID controller are K 共 ␭⫹ ␪ 兲
␭2 ␪
T F⫽ , T I ⫽ ⫹ ␶ ⫺T F , Let C ( s ) ⫽ ( ␶ s⫹1 ) / ( ␭s⫹1⫺e ⫺ ␪ s ) ⫽ f ( s ) /s.
2␭⫹ ␪ /2 2 The controller can be expanded in a Maclaurin
series as
␪␶ TI

冋 册
T D⫽ ⫺T F , K C⫽ .
2T I K 共 2␭⫹ ␪ /2兲 1 f ⬙共 0 兲 2
C共 s 兲⫽ f 共 0 兲 ⫹ f ⬘ 共 0 兲 s⫹ s ⫹¯ .
Since Eq. 共13兲 is used in many papers, for com- s 2!
parison, only this form is considered in the later 共18兲
part of this paper.
Taking the first three terms, we obtain the Maclau-
If a conventional PID controller is installed, the
rin PID controller, of which the parameters are 关9兴
parameter T F has been determined. Usually T F
⫽0.1T D 关4兴. In this case, one can also use the
␪2
above rules by simply omitting the new T F . Then, T F ⫽0, T I⫽ ⫹␶,
similar results will be obtained. 2 共 ␭⫹ ␪ 兲

冋 册
Both the H 2 PID controller and Maclaurin PID
controller are based on the result of IMC. In IMC, ␪2 ␪ TI
T D⫽ 1⫺ , K C⫽ .
the plant is factored as 2 共 ␭⫹ ␪ 兲 3T I K 共 ␭⫹ ␪ 兲
G 共 s 兲 ⫽G ⫹ 共 s 兲 G ⫺ 共 s 兲 , Sometimes, one may use a second-order model.
where G ⫹ ( s ) contains all nonminimum phase fac- The above methods can be directly extended to
tors, G ⫹ ( 0 ) ⫽1, and G ⫺ ( s ) is the minimum phase this case. Since the design procedure is almost the
portion of the model. The IMC controller is then same, only the results are given here. Suppose that
given by the following formula 关8兴: the model is

1 Ke ⫺ ␪ s
Q共 s 兲⫽ , G共 s 兲⫽ . 共19兲
共 ␭s⫹1 兲 n G ⫺ 共 s 兲 共 ␶ 1 s⫹1 兲共 ␶ 2 s⫹1 兲
where n is chosen so that the controller is bi- Utilizing the Taylor series, the PID controller pa-
proper, that is, both Q ( s ) and 1/Q ( s ) are proper. rameters designed by the proposed method are
For the given plant we have
␭2 ␶ 1␶ 2
␶ s⫹1 T F⫽ , T I⫽ ␶ 1⫹ ␶ 2 , T D⫽ ,
C共 s 兲⫽ , 共16兲 2␭⫹ ␪ TI
␭s⫹1⫺e ⫺ ␪ s
Since a time delay is included in the controller, TI
K C⫽ . 共20兲
it cannot be directly implemented. Many methods K 共 2␭⫹ ␪ 兲
have been presented to approximate the control-
ler by a rational transfer function, such as a nu- The H 2 PID controller parameters are
merical method, Padé approximation, Taylor se-
ries, Maclaurin series, and so on 关11兴. The H 2 sub- ␭␪ ␶ 1␶ 2
optimal PID controller given by 关7兴 utilizes a Padé T F⫽ , T I⫽ ,
␭⫹2 ␪ ␶ 1⫹ ␶ 2
approximation. The resultant controller is
1 共 ␶ s⫹1 兲共 1⫹ ␪ s/2兲 TI
C共 s 兲⫽ 共17兲 T D⫽ ␶ 1⫹ ␶ 2 , K C⫽ . 共21兲
Ks ␭ ␪ s/2⫹␭⫹ ␪ K 共 ␭⫹2 ␪ 兲

and the PID controller parameters are The Maclaurin PID controller parameters are
 Weidong Zhang et al. / ISA Transactions 41 (2002) 511–520 515

2␭ 2 ⫺ ␪ 2 and
T F ⫽0, T I⫽ ␶ 1⫹ ␶ 2⫺ ,
2 共 2␭⫹ ␪ 兲 1⫺0.5s
T共 s 兲⫽ .
2␭ 2 ⫺ ␪ 2 ␶ 1 ␶ 2 ⫺ ␪ 3 / 共 12␭⫹6 ␪ 兲 共 ␭s⫹1 兲共 1⫹0.5s 兲
T D ⫽⫺ ⫹ ,
2 共 2␭⫹ ␪ 兲 TI Only one pole of the nominal plant is canceled.
Note that there is an adjustable parameter ␭ in
TI the three PID controllers. It has been shown by
K C⫽ . 共22兲
K 共 2␭⫹ ␪ 兲 some researchers that ␭ relates directly to the
nominal performance and robustness of the sys-
3. Discussion tem. This paper will show that the quantitative
time-domain response and frequency-domain re-
All three PID controllers are suboptimal. The sponse can be gained by adjusting ␭.
design procedure shows that they relate closely Consider the H ⬁ PID controller again. We re-
and each has its own features. The difference be- gard the error introduced by the approximation as
tween the Maclaurin PID controller and the H 2 uncertainty. The actual plant is in the form of a
and H ⬁ PID controllers is obvious. The difference first-order plus time delay. Then
between the H 2 PID controller and the H ⬁ PID
共 1⫹ ␪ s/2兲 e ⫺ ␪ s
controller lies in the manner they cancel the poles T共 s 兲⫽
of the plant. The proposed H ⬁ controller tends to ␭ 2 s 2 ⫹ 共 2␭⫹ ␪ /2兲 s⫹ 共 1⫹ ␪ s/2兲 e ⫺ ␪ s
cancel all poles of the process while the H 2 con- 共23兲
troller tends to cancel the poles of the minimum and
phase part of the process.
␭ 2 s 2 ⫹ 共 2␭⫹ ␪ /2兲 s
S共 s 兲⫽ .
3.1. Example 1 ␭ 2 s 2 ⫹ 共 2␭⫹ ␪ /2兲 s⫹ 共 1⫹ ␪ s/2兲 e ⫺ ␪ s
共24兲
The difference between the H 2 PID controller
and the H ⬁ PID controller is illustrated in this ex- It is seen that T ( s ) does not depend on K and ␶.
ample. Consider the plant described by This implies that the setpoint response for the
closed-loop system relates only to ␭ and ␪. The
e ⫺s similar result also exists for the H 2 PID controller.
G共 s 兲⫽ . Strictly speaking, the setpoint response of the sys-
s⫹1
tem with the Maclaurin PID controller relates to
The nominal plant with Padé approximation is not only ␭ and ␪ but ␶. Since the difference is very
1⫺0.5s
G共 s 兲⫽ .
共 s⫹1 兲共 1⫹0.5s 兲
The H ⬁ method yields that
共 s⫹1 兲共 1⫹0.5s 兲
Q共 s 兲⫽
共 ␭s⫹1 兲 2
and the nominal complementary sensitivity trans-
fer function is
1⫺0.5s
T共 s 兲⫽ .
共 ␭s⫹1 兲 2
We see that both of the two poles of the nominal
plant are canceled. For the H 2 controller we have
s⫹1 Fig. 3. Responses of system A. 共Solid line: H ⬁ ; dashed
Q共 s 兲⫽
␭s⫹1 line: Maclaurin; dotted line: H 2 .兲
516 Weidong Zhang et al. / ISA Transactions 41 (2002) 511–520

Fig. 4. Responses of system B. 共Solid line: H ⬁ ; dashed Fig. 6. Quantitative rise time. 共Solid line: H ⬁ ; dashed line:
line: Maclaurin; dotted line: H 2 .兲 Maclaurin; dotted line: H 2 .兲

small, it can be considered that its setpoint re-

sponse relates only to ␭ and ␪. overshoot. The resultant parameters are ␭⫽0.41␪
for the H 2 controller, ␭⫽0.46␪ for the H ⬁ PID
3.2. Example 2 controller, and ␭⫽0.42␪ for the Maclaurin PID
controller. The closed-loop responses are shown in
The three suboptimal PID controllers are com- Fig. 3 and Fig. 4. For the two plants, the H 2 con-
pared in this example. Consider the following two troller and the H ⬁ PID controller have almost the
plants: same response. For the plant with a small ratio of
␪ and ␶, A, the H 2 controller and the H ⬁ PID
e ⫺0.5s e ⫺5s controller have faster rise times and settling times
A: G 共 s 兲 ⫽ , B: G 共 s 兲 ⫽ .
s⫹1 s⫹1 than that of the Maclaurin PID controller, and for
the plant with a large ratio of ␪ and ␶, B, the H 2
A unit step setpoint is added at t⫽0s and a unit controller and the H ⬁ PID controller have a slower
step load disturbance at t⫽6s or 60s. The param- settling time and disturbance rejection.
eter in all the three controllers is adjustable. To It is found that the response shapes of the H 2
obtain a fair comparison, the parameter is adjusted controller and the H ⬁ PID controller for the two
such that the closed-loop responses have the same plants are almost the same. This validates the

Fig. 5. Quantitative overshoot. 共Solid line: H ⬁ ; dashed Fig. 7. Quantitative ISE. 共Solid line: H ⬁ ; dashed line:
line: Maclaurin; dotted line: H 2 .兲 Maclaurin; dotted line: H 2 .兲
 Weidong Zhang et al. / ISA Transactions 41 (2002) 511–520
above analysis. We define ␭ as the ‘‘performance
degree.’’ It is seen that the overshoot and the ratio
of rise time and ␪ are determined only by the ratio
of the performance degree and ␪ 共Figs. 5 and 6兲.
The break in Fig. 6 is caused by the difference of
the definition of the rise time for systems with
overshoot and without overshoot. The empirical
formulas for estimating the two indices can also be
given. For example, the formulas of the H ⬁ PID
controller are as follows:
Overshoot
再
⫺0.86共 ␭/ ␪ 兲 3 ⫹14.21共 ␭/ ␪ 兲 2
⫽ ⫺8.72␭/ ␪ ⫹1.86, 0.1⭐␭/ ␪ ⭐0.59
0, 0.59⭐␭/ ␪ ⭐1.2,
共25兲
Rise time
再
30.34共 ␭/ ␪ 兲 3 ⫺24.63共 ␭/ ␪ 兲 2
⫽ ⫹8.48␭/ ␪ ⫺0.45, 0.1⭐␭/ ␪ ⭐0.59
3.97␭/ ␪ ⫺1.02, 0.59⭐␭/ ␪ ⭐1.2.
共26兲
In 关12,13兴, nonovershoot and monotone nonde-
creasing response is studied. For the three control-
lers it is very easy to get such a response.
Both the H 2 control and the H ⬁ control relate to
the integral square error 共ISE兲. The H 2 control
minimizes the ISE for a particular input and the
H ⬁ control minimizes the worst ISE resulting
from any two-norm bounded inputs. It is found
Fig. 8. Quantitative resonance peak. 共Solid line: H ⬁ ;
dashed line: Maclaurin; dotted line: H 2 .兲
517
that the ratio of ISE and ␪ is also determined by
the performance degree 共Fig. 7兲. This implies that
the relationship between the classical performance
indices and the optimal performance indices is es-
tablished.
In frequency-domain analysis, one important
concept is the resonance peak, that is, the maxi-
mum modulus of the closed-loop transfer function.
In 关14兴, it is referred to as the maximum log
modulus and utilized to design PID controller. A
commonly used specification for it is 2dB. The
quantitative relationship between the resonance
peak and the performance degree is shown in Fig.
8.
Another important concept in frequency domain
is stability margin, on which numerous methods
are based 共see, for example, 关15兴兲. The magnitude
stability margin and phase stability margin provide
intuitive tools for control system design and are
very familiar to engineers. The quantitative rela-
tionship between the stability margin and the per-
formance degree is shown in Figs. 9 and 10.
Recently, 关16兴 proposed a new frequency-
domain index for PID controller design,
1/max兩 Re关 G 共 j ␻ 兲 C 共 j ␻ 兲兴 兩 ,
which involves both the magnitude stability mar-
gin and phase stability margin to a certain extent.
It is found that there also exists the quantitative
relationship between the new index and the perfor-
mance degree 共Fig. 11兲.
Unfortunately, the simple quantitative relation-
ship comes into existence only for setpoint re-
sponse. The transfer function from the load distur-
Fig. 9. Quantitative magnitude margin. 共Solid line: H ⬁ ;
dashed line: Maclaurin; dotted line: H 2 .兲
518 Weidong Zhang et al. / ISA Transactions 41 (2002) 511–520
Fig. 10. Quantitative phase margin. 共Solid line: H ⬁ ;
dashed line: Maclaurin; dotted line: H 2 .兲
bance to the system output is G ( s ) S ( s ) . Thus, the
disturbance response relates not only to ␭ and ␪,
but K and ␶. We define the perturbance peak to be
the maximum output caused by the disturbance
关10兴. For the H ⬁ controller the relationship is
shown in Fig. 12.
For the second-order model, the closed-loop re-
sponse relates both ␶ 1 and ␶ 2 . If one of them is
determined, a similar quantitative relationship also
exists, but not so clearly.
There always exists uncertainty in practice. Sup-
pose that the norm-bounded uncertainty is de-
scribed by
冏 冏
G̃ 共 s 兲 ⫺G 共 s 兲
⫽ 兩 ⌬ m 共 j ␻ 兲 兩 ⭐⌬ 共 ␻ 兲 ,
G共 s 兲
Fig. 11. Quantitative new index. 共Solid line: H⬁ ; dashed
line: Maclaurin; dotted line: H2 .兲
Fig. 12. Quantitative perturbance peak.
where ⌬ ( ␻ ) is the uncertainty profile. Let S ( s )
and T ( s ) be the sensitivity transfer function and
the complementary sensitivity transfer function of
the approximated system, respectively. The robust
stability is met if and only if
储 ⌬ 共 ␻ 兲 T 共 j ␻ 兲 储 ⬁ ⬍1, ᭙␻. 共28兲
This implies that arbitrary robust stability can be
gained by increasing the performance degree ␭.
Assume that the performance is 储 W ( s ) S ( s ) 储 ⬁
⬍5 ␪ . Then, the robust performance is met if and
only if
储 兩 W 共 j ␻ 兲 S 共 j ␻ 兲 /5␪ 兩 ⫹ 兩 ⌬ 共 ␻ 兲 T 共 j ␻ 兲 兩 储 ⬁ ⬍1, ᭙␻.
共29兲
The quantitative uncertainty profile that guaran-
tees robust performance for an H ⬁ PID controller
共27兲 is shown in Fig. 13.
In classical control theory, the controller is usu-
ally designed for the nominal performance speci-
fication and used for the control of the uncertain
plant. If a good model can be obtained, the esti-
Fig. 13. Quantitative uncertainty profile.
 Weidong Zhang et al. / ISA Transactions 41 (2002) 511–520
mates given by the above study are in good agree-
ment with the actual time-domain responses. If the
uncertainty scope is large, one may hope to know
whether the required performance is obtained or
not. Eqs. 共28兲 and 共29兲 give a perfect estimate.
However, they are more mathematical than practi-
cal and very inconvenient. As a matter of fact, we
usually wish to estimate the ‘‘worst case’’ re-
sponse 共i.e., the gain and time delay are at their
upper limits, while the time constant is at its lower
limit兲 or the range of the closed-loop response.
关17兴 recently developed a method for designing
IMC systems with parametric uncertainty. First, a
specified value M p is chosen based on the maxi-
mum desired overshoot, or the ‘‘worst case’’ over-
shoot. Second, the magnitude of the complemen-
tary sensitivity transfer function is designed to be
equal to or less than M p. For example, for the
‘‘worst case’’ overshoot 10%, M p⫽1.05. By solv-
ing 兩 T ( j ␻ ) 兩 ⭐M p, a unique solution on ␭ can be
obtained.
Here, an alternative method is proposed to esti-
mate the ‘‘worst case’’ response or the range of the
closed-loop response of the proposed controller.
Assume that we hope to achieve the ‘‘worst case’’
overshoot of 5%. The new procedure is as follows:
共i兲 Design the controller for the nominal
plant. For 5% overshoot, ␭⫽0.5␪.
共ii兲 Substitute the nominal plant by the
⫹
‘‘worst case’’ plant K ⫹ e ⫺ ␪ s / ( ␶ ⫺ s⫹1 ) .
共iii兲 Increase the performance degree ␭
monotonically until the overshoot is
equal to 5%.
For system with time delay, the typical step in
共iii兲 is 0.01␪, and for systm without time delay, the
typical step in 共iii兲 is 0.01␪. This procedure can
easily be performed in Matlab with SIMULINK.
4. Conclusions
In this paper, an efficient H ⬁ optimal method for
designing a PID controller was developed. Ex-
plicit formulas are derived analytically. The result
is similar to the H 2 PID controller, but different in
the manner of canceling the poles of the plant.
The most important contribution of this paper is
that the concept ‘‘performance degree’’ is defined
and the quantitative relationship between the
closed-loop response and the performance degree
is given. In systems with the H 2 PID controller or
519
the H ⬁ PID controller, the relationship is exact,
and for the Maclaurin PID controller, the relation-
ship roughly exists. Thus, all three controllers can
be designed for desired time-domain response or
frequency-domain response. The given quantita-
tive relationship between the classical perfor-
mance indices and the optimal performance indi-
ces makes it possible to build a bridge between
the classical design method and modern design
methods.
For practical purposes, the H 2 PID controller
may be the most convenient one, because it has a
linear relationship with many performance indices.
Since there always exists uncertainty in practice,
the quantitative design for the system with para-
metric uncertainty is also discussed, and a new
design method is proposed to estimate the ‘‘worst
case’’ response or the range of the closed-loop re-
sponse of the proposed controller.
It is also shown that the suboptimal PID gives a
satisfactory approximation to the exact time delay
compensated scheme. This implies that the PID
controller can provide relatively good response for
systems with time delay, even when the time delay
is very large.
Acknowledgments
This project was supported by the National
Natural Science Foundation of China 共Grant
No. 69804007兲 and the National Key Technologies
R&D Program in the Tenth Five-Year Plan 共Grant
No. 2001BA201A04兲.
References
关1兴 Ziegler, J. G. and Nichols, N. B., Optimum settings for
automatic controllers. Trans. ASME 64, 759 共1942兲.
关2兴 Cohen, G. H. and Coon, G. A., Theoretical consider-
ations of retarded control. Trans. ASME 75, 827
共1953兲.
关3兴 Lopez, A. M., Smith, C. L., and Murrill, P. W., Con-
troller tuning relationships based on integral perfor-
mance criteria. Instrum. Technol. 14, 57 共1967兲.
关4兴 Astrom, K. J. and Hagglund, T., PID Controllers:
theory, design, and tuning, 2nd ed. ISA, NC, 1995.
关5兴 Sung, S. W., Lee, J., and Lee, I., A new tuning rule and
modified PID controller. Ind. Eng. Chem. Res. 34,
4127 共1995兲.
关6兴 Astrom, K. J., Panagopoulos, H., and Hagglund, T.,
Design of PI controllers based on nonconvex optimi-
zation. Automatica 34 共5兲, 585 共1998兲.
关7兴 Rivera, D. E., Morari, M., and Skogestad, S., Internal
model control: 4-PID controller design. Ind. Eng.
Chem. Res. 25, 252 共1986兲.
520 Weidong Zhang et al. / ISA Transactions 41 (2002) 511–520

关8兴 Morari, M. and Zafiriou E., Robust Process Control.

Prentice-Hall, Englewood Cliffs, NJ, 1989. Yugeng Xi was born in the
People’s Republic of China in
关9兴 Lee, Y., Park, S., Lee, M., and Brosilow, C., PID con- 1946. He graduated from
troller tuning for desired closed loop response for Haerbin Engineering College
SISO systems. AIChE J. 44 共1兲, 106 共1998兲. in 1968. From 1979 to 1984 he
was a visiting scholar at the In-
关10兴 Zhang, W. D., Xu, X. M., and Sun, Y. X., Quantitative stitute of Control Technology,
performance design for integrating processes with Technical University Munich,
time delay. Automatica 35 共4兲, 719 共1999兲. Germany, where he received
关11兴 Zhang, W. D., Robust control of systems with time the Dr.-Ing. degree in 1984.
Since then, he has been work-
delay, Ph.D thesis, Zhejiang University, 1996. ing at Shanghai Jiaotong Uni-
关12兴 Lin, S. K. and Fang, C. J., Nonovershoot and mono- versity. He received awards for
tone nondecreasing step response of a third order Achievement in Science and
SISO linear system. IEEE Trans. Autom. Control 42 Technology six times by the State Commission of Education and the
Shanghai government. Prof. Xi was the vice chair of the IFAC TC
共9兲, 1299 共1997兲. Large Scale Systems and is now the vice president of the Chinese
关13兴 Dan-Isa, A. and Atherton, D. P., Time-domain method Association of Automation. His current research interests include pre-
for the design of optimal linear controllers. IEE Proc.: dictive control, robust control, robot control, and computer vision.
Control Theory Appl. 144, 287 共1997兲.
关14兴 Luyben, W. L., Simple method for tuning SISO con-
trollers in multivariable systems. Ind. Eng. Chem. Pro- Genke Yang was born in the
People’s Republic of China in
cess Des. Dev. 25, 654 共1986兲. 1963. He received his B.Sc.
关15兴 Ho, W. K., Lim, K. W., Hang, C. C., and Ni, L. Y., and M.Sc. degrees in math-
Getting more phase margin and performance out of ematics, and Ph.D. degree in
systems engineering. He is cur-
PID controllers. Automatica 35, 1579 共1999兲.
rently an associate professor in
关16兴 Wang, Y. G. and Shao, H. H., Optimal tuning for PI the department of automation,
controllers. Automatica 36 共1兲, 147 共2000兲. Shanghai Jiaotong University.
关17兴 Stryczek, K., Laiseca, M., Brosilow, C., and Leitman, He is the author of 40 papers.
His current research interests
M., Tuning and design of single-input, single-output include modeling, analysis,
control systems for parametric uncertainty. AIChE J. and control theory of discrete
46 共8兲, 1616 共2000兲. event dynamic systems and hy-
brid dynamic systems.

Xiaoming Xu was born in the

Weidong Zhang was born in
the People’s Republic of China
in 1967. He received the BS,
MS, and Ph.D degree from
Zhejiang University in 1990,
1993, and 1996, respectively.
He worked in National Key
Laboratory of Industrial Con-
trol Technology as a postdoc-
toral research fellow before
joining Shanghai Jiaotong Uni-
versity in 1998 as an associate
professor. Since 1999 he has
been a professor at the Depart-
ment of Automation, Shanghai Jiaotong University. He is the author of
68 papers. His current research interests include process control, ro-
bust control, field-bus, and digital signal processing.
People’s Republic of China in
1957. He received the BS de-
gree from Huazhong Univer-
sity of Science and Technology
in 1982, and the MS and Ph.D
degrees from Shanghai Jiao-
tong University in 1984 and
1987, respectively. From 1988
to 1990 he worked in Germany
as an Alexander von Humboldt
research fellow. He joined
Shanghai Jiaotong University
in 1990 as a professor in the
department of automation. Since 1997 he has been the deputy princi-
pal of Shanghai Jiaotong University. He is the head of numerous state
research projects and the author of numerous papers. His current re-
search interests include control theory, artificial intelligence, computer
networking, and digital signal processing.