Chapter 11 – Dynamic Behavior

and Stability of Closed-Loop

Control Systems
CHEN 4130

Learning with Purpose

 Agenda
• Feedback Control System
• Block Diagram Representation
• Closed-Loop Transfer Functions
– Set Point Changes
– Disturbance Changes
• Closed-Loop Responses of Simple Control Systems
– Proportional (P) Control
– Proportional-Integral (PI) Control
• Stability of Closed-Loop Systems
• Root Locus Diagram

2
 Feedback Control System

Operations performed by a feedback control

system:
– Measures process output variable using a
sensor system.
– Measured value of output variable is sent to
controller and compared to setpoint.
– Controller determines appropriate action and
sends a signal to final control element.
– Final control element uses controller signal to
adjust a process manipulated variable.

3
 Block Diagram Representation
Example – Surge Tank
Consider a surge tank process in which the feed flow rate (q1) is a
disturbance variable D, the exit flow rate (q) is the manipulated
variable U and the height (h) is the controlled variable Y
q1 hsp
h hm
LT LC

h p

I/P

pT
q

Other notation above: hm is the measured height, hsp is the level

setpoint, p is the controller output, and pT is the signal to the valve.
4
Next, we develop a transfer function for each of the five elements in the
feedback control loop. For simplicity, assume the starting point is at steady-
state.

#1 - Process
dh
The Surge Tank Mass Balance gives: A = q1 – q
dt
Converting to Laplace form using deviation variables gives the following
transfer function models for the manipulated and disturbance variables:

H '(s) = G p (s)Q '(s) + G d (s)Q1' (s)

K1 − 1 / A K 2 1/ A
where: G p (s) = = G d (s) = =
s s s s

Hu’(s)
Q’(s) Gp(s) = K1/s

Hd’(s) +
Q1’(s) Gd(s) = K2/s + H’(s)
5
#2 - Composition Sensor-Transmitter (Analyzer)
We assume that the dynamic behavior of the composition sensor-
transmitter can be approximated by a first-order transfer function.

Km
H 'm (s) = G m (s)H ' (s) where: G m (s) =
ms + 1

H’(s) Gm(s) = Km/(τms + 1) Hm’(s)

6
#3 - Controller
We assume that an electronic PID controller is used, acting on the error
between the setpoint and the level measurement. The transfer function for
a PID controller is given by:
 1 
P (s) = G c (s)E(s)
'
where: Gc ( s ) = K c 1 + +  D s 
 Is 

 1 
E(s) G c (s) = K c 1 + +  D s  P’(s)
 Is 

where: Kc τI τD are controller tuning parameters

E(s) is the Laplace transform of the error signal Hsp’(s) – Hm’(s)

P’(s) is the Laplace transform of the controller output

7
#4 - Current-to-Pressure (I/P) Transducer
The transducer transfer function merely consists of a steady-state gain KIP:

Pt' (s) = K IP P ' (s)

P’(s) KIP PT’(s)

𝒑𝒏𝒆𝒖𝒎𝒂𝒕𝒊𝒄 𝒔𝒊𝒈𝒏𝒂𝒍 𝒐𝒖𝒕

KIP =
𝒆𝒍𝒆𝒄𝒕𝒓𝒐𝒏𝒊𝒄 𝒔𝒊𝒈𝒏𝒂𝒍 𝒊𝒏

15 −3 𝑝𝑠𝑖
= = 0.75 psi/mA
20−4 𝑚𝐴

8
#5 - Control Valve
Control valves are usually designed so that the flow rate through the valve is
a nearly linear function of the signal to the valve actuator. Therefore, a first-
order transfer function is an adequate model.

Kv
Q ' (s) = G v (s)Pt' (s) where: G v (s) =
 vs + 1

PT’(s) Gv(s) = Kv/(τvs + 1) Q’(s)

9
 Block Diagram Representation
The overall closed loop response for the surge tank problem
can be represented by the following block diagram.
Q1 (s) H d (s)
G d (s)

 (s )
H sp E (s) P (s) Pt (s) Q (s) H u (s ) H (s)
G c (s) K IP G v (s) G p (s)

H m (s)
G m (s)

Note the use of the following block flowsheet items:

1. Summer

2. Comparator

3. Block  Y(s) = G(s)X(s)

10
 Closed-Loop Transfer Function
The standard block diagram for a SISO control system is given
below:
D
Gd
Yd
Ysp ~ E P U Yu Y
Ysp +
Km +- Gc Gv Gp +

Ym
Gm

The closed-loop transfer function model for a simple feedback

control system can be derived from the above diagram, and is
given by:
K mG cG vG p Gd
Y= Ysp + D
1 + G cG vG pG m 1 + G cG vG pG m

11
Closed-loop transfer function for a set point change, D = 0:

Y K mGc Gv G p
=
Ysp 1 + Gc Gv G p Gm

Closed-loop transfer function for a disturbance change, Ysp = 0:

Y Gd
=
D 1 + Gc Gv G p Gm

12
 General Stability Criteria
• Most industrial processes are stable without feedback
control. Thus, they are said to be open-loop stable or self-
regulating.
• An open-loop stable process will return to the original
steady state after a transient disturbance (one that is not
sustained) occurs.
• By contrast there are a few processes, such as exothermic
chemical reactors, that can be open-loop unstable.

Definition of Stability. An unconstrained linear system is

said to be stable if the output response is bounded for all
bounded inputs. Otherwise it is said to be unstable.

13
 Stability of Closed-Loop Systems
• A linear closed-loop dynamic system is stable if the output
is bounded for a bounded input.
• Stability can be determined by examining the roots of the
characteristic equation (poles of system transfer function):

Y K mG cG vG p Y Gd
= =
Ysp 1 + G c G v G p G m D 1 + G cG vG pG m

• Let GOL = GcGvGpGm so that 1 + GcGvGpGm = 1 + GOL

• A control system is stable if all poles of the characteristic
equation 1 + GOL = 0 have negative real parts.
• The poles represent the natural response modes of the
control system. 14
Chapter 11 General Stability Criteria

Stability regions in the

complex plane for roots
of the characteristic
equation.
 Roots of 1 + GcGvGpGm

(Note complex roots

always occur in pairs)

Figure 11.26 Contributions of characteristic equation roots to closed-loop response.

 Stability of Closed-Loop Systems
• Some methods for investigating the stability of linear
control systems:

➢ Solve for roots of characteristic equation

➢ Routh stability test

➢ Direct substitution method

➢ Frequency response methods

17
 Stability of Closed-Loop Systems
Method 1 – Solve for roots of characteristic equation

• Analytical methods can be used for first or second order

characteristic equations.
• For higher order equations, numerical methods are used
(e.g., root solving routine in Matlab).

Example: Characteristic Equation: s2 + 3s + (2+Kc) = 0

Solving for the roots of this equation gives:

s = - 3Τ2 ± 1 − 4𝐾𝑐 / 2

System is stable as long as Kc > -2

18
 Stability of Closed-Loop Systems
Method 2 – Routh Stability Test

• Write characteristic equation in polynomial form:

a n s n + a n −1s n −1 + a n − 2s n − 2 + ... + a 1s + a o = 0

• Two conditions that must be satisfied before continuing

with the test:
➢ There must be (n + 1) terms in characteristic equation
➢ All coefficients must be positive.

19
 Stability of Closed-Loop Systems
Method 2 – Routh Stability Test (continued)

• For higher order systems (n > 2), the Routh Array is

constructed
a n s n + a n −1s n −1 + a n − 2s n − 2 + ... + a 1s + a o = 0

Routh Array Row

1 an an-2 an-4 …
a a − a n a n −3
b1 = n −1 n − 2 2 an-1 an-3 an-5 …
a n −1
3 b1 b2 b3
a n −1a n − 4 − a n a n −5 4 c1 c2 …
b2 =
a n −1 … … …
n+1 z1
b1a n −3 − a n −1b 2
c1 =
b1

ba − a n −1b 3
Stability Criteria - A system is stable if all
c 2 = 1 n −5 elements in the first column are positive.
b1
20
 Stability of Closed-Loop Systems
Method 3 – Direct Substitution Method
• Set s = j in the characteristic equation
• Solve for Kcm and c
(a) one equation for real part
(b) one equation for imaginary part

21
 Stability of Closed-Loop Systems
Method 3 – Direct Substitution Method (continued)

Example: Characteristic equation is 8s 3 + 14s 2 + 7s + (1 + K c ) = 0

Let s = j: → (1 + K c ) ( )
− 142 + 7 − 83 j = 0
Real part: (1 + K − 14 ) = 0
c
2

Imaginary part: (7 − 8 ) = 0

2

Solving for ω in the imaginary part equation and substituting

into the real equation gives allowed values for Kc:

For  = 0, Kc = −1 Controller must be reverse

acting with Kc > 0.
7
For  = , Kc = 11.25
2
8 For stability, 0 < Kc < 11.25.
22