Sunteți pe pagina 1din 38

Process Modeling & Control

Books:
1. Process Control, Modeling, Design and Simulations,
by B. Wayne Bequette
2. Process Dynamics, Modeling and Control,
by B. A. Ogunnaike, and W. H. Ray
3. Process Control (Designing Processes and Control Systems for
Dynamic Performance), T. E. Marlin
4.

Principles and Practice of Automatic Process Control, C. A. Smith

and A. B. Corripio

Dr. Muhammad Zaman, SE, PIEAS

Introduction

Dr. Muhammad Zaman, SE, PIEAS

Introduction
Control Objectives
Input variables (manipulated or disturbance variables)
Output variables (measured or unmeasured)
Constraints (hard or soft)
Operating characteristics (continuous or batch or semibatch
(semicontinuous))
Safety, environment and economic considerations
Control structure (feedback or feedforward)

Dr. Muhammad Zaman, SE, PIEAS

Introduction
Figure 1.1: Tank level problem / Surge tank

Dr. Muhammad Zaman, SE, PIEAS

Introduction
Feedback Control (Scenario 1)
Process 2 regulated F2, depending on steam demand
F2 is disturbance stream, as regulation of F2 is by another
system
Use F1 as manipulated variable
F1 is adjusted to maintain a desired tank height

Dr. Muhammad Zaman, SE, PIEAS

Introduction
Feedback Control (Scenario 1)

Dr. Muhammad Zaman, SE, PIEAS

Introduction
Feedback Control (Scenario 2)
Process 1 regulates the flow rate F1, disturbance
Adjust F2 to maintain tank height, manipulated variable
Control valve, fail-open or air-to-close

Dr. Muhammad Zaman, SE, PIEAS

Introduction

Dr. Muhammad Zaman, SE, PIEAS

Introduction
Feedforward Control

Dr. Muhammad Zaman, SE, PIEAS

Introduction
Feed-forward/ feedback control structure

Dr. Muhammad Zaman, SE, PIEAS

Introduction
Tune level controller for fast or slow response?
Preferable to tune slow return to the set-point
Mostly true for scenario 2
Outlet flow is manipulated but affect process 2
Change the outlet flowrate slowly, yet fast enough that the
tank does not overflow or go dry
Importance of the dominant timescale of the process e.g.,

Dr. Muhammad Zaman, SE, PIEAS

Introduction
Taking a Shower
Multivariable control
Control objectives
Input variables
Output variables
Constraints
Operating characteristics
safety, environment and economic consideration
Control structure (FB or FF)

Dr. Muhammad Zaman, SE, PIEAS

Introduction
Instrumentation
Sensor
Actuator
Controller
Continuous or discrete signals
Analog
Current, voltage and pneumatic
Digital
D/I, I/P

Dr. Muhammad Zaman, SE, PIEAS

Introduction
Instrumentation
Control valve placement

Dr. Muhammad Zaman, SE, PIEAS

Introduction
Process Model and Dynamic Behavior
First principle based
Imperial models
Input vs output relations

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control

What is control algorithm?


How does the controller change the flow rate to the
process?
Number of possible algorithms
Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


On-off controller

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


Proportional Control
Make the signal to valve proportional to the error

b is the bias term and kc is the proportional gain

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


Process gain is simply SS change in output for a SS
change in process input
Process gain
Valve gain
Overall gain (valve and process)

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


For the second strategy, the process gain is negative (an
increase in F2 causes a decrease in h)

Valve gain

The gain between valve-top pressure and tank height is

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


Valve Gains
3 and 15 psig
Fail close
Maximum flow rate through the valve is 120 gpm

Similarly, a fail-open valve of the same size would have a


negative gain (-10 gpm/psig)
Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


Development of Control Block Diagrams
All dynamic elements in a control loop are combined, using
their Laplace transfer function representation
Block diagrams are based on Laplace domain signals,
which are assumed to be in deviation variable (perturbation
from steady state) form

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


The block diagram for previous P&ID has the following
components
Level controller (relates error to controller output)
Valve (relates controller output signal to flow-through valve)
Process (relates manipulated input to process output)
Disturbance (relates the disturbance input to the process
output)
Sensor (measures tank level)

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


Controller Transfer Function
Compares the hm to the hsp.
From the proportional control law (algorithm) for this system
b = Pv1s (SS pressure to valve), we find the controller inputoutput relationship

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


Where the controller input signal is the error
We write the equation [where, c(s) is the controller
output, the pressure to the valve]
Transfer function form for a proportional-only controller

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


When analyzing block diagrams, the comparator is shown
outside the controller transfer function block.
The block diagram uses r(s) to represent the set-point
The orientation for the TF representation of the controller
is slightly different from that of the physical system

Block diagram relationship for controller transfer function.


Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


Valve transfer function
The input signal is the pressure to the valve top and the
output is the flow rate of fluid through the valve

Block diagram of the valve and physical system


Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


Process transfer function
The input to the block is the flow rate to the tank, and the
output is the tank level.
Block diagram of the manipulated input effect on the
process

Block diagram of the manipulated input effect on the process


Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


Disturbance transfer function
The disturbance input is the flow rate from the tank and the
output is the tank level
Block diagram of the disturbance input effect on the process

Block diagram of the disturbance input effect on the process


Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


Measurement (Sensor) transfer function
The input to the block is the tank level and the output is the
actual measurement of tank level.

Block diagram of sensor


Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


Control block diagram
Combine the previous five figures to obtain the feedback
control system block diagram shown below.
Two externally supplied signals, the set-point and the
disturbance

Control system block diagram

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


We often assume that the output variable can be perfectly
measured, and that process input (usually a flow rate) is
directly manipulated; in this case we do not include the
valve and measurement TFs in the closed-loop block
diagram
Equivalently, we can lump the valve and measurement
dynamics into the process TF, again allowing us to neglect
the valve and measurement TFs.
Similarly, the measurement device can be lumped into the
disturbance transfer function

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


Response to set-point changes
assuming gm(s)=gv(s)=1
This simplification is equivalent to lumping the
measurement and valve dynamics into the process TF.
Notice that we are focusing on set-point changes only, so
we have not included the disturbance block.

Fig. 5.10
Simplified control block diagram, disturbances are neglected, valve &
measurement dynamics are lumped into the process TF
Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


To find the output response to a set-point change
A critical aspect is to determine the closed-loop stability
Use block diagram manipulation to find the relationship
between the set-point and the output
Closed-loop diagram of previous figure into a single TF
block, to provide the closed-loop relationship between setpoint and process output.

Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


To obtain the output y(s) as a function of set-point, r(s)
Combine the process input-output relationship, y(s) = gp(s)
u(s), with controller relationship, u(s) = gc(s) e(s), to find
The error is defined as e(s) = r(s) y(s), we can write

Relationship between r(s) & y(s) as the closed-loop


transfer function, gCL(s),
Dr. Muhammad Zaman, SE, PIEAS

Introduction to Feedback Control


If all of the poles of gCL(s) are stable, then the closed-loop
system is stable
The denominator of gCL(s) is also known as the
characteristic equation
Previous equation has been derived for the simple
diagram shown in fig. 5.10.
Please realize that more-complex block diagrams will have
more-complex closed-loop transfer functions.
The closed-loop TF for figure 5.9 (neglecting disturbance)

Dr. Muhammad Zaman, SE, PIEAS

S-ar putea să vă placă și