H As Se N: Feedback Control System Characteristics and Performance
H As Se N: Feedback Control System Characteristics and Performance
se
S Z Sayed Hassen
as
Contents
1 Introduction 1
3 Sensitivity
4 Transient Response
H 2
6
d
5 Steady-state error 8
6 Example 9
ye
1 Introduction
Introduction
The use of the error signal e to control the process results in a closed-loop
Sa
open-loop system: The disturbance Td (s) directly influences the output Y (s). In
1
the absence of feedback, the control system is highly sensitive to disturbances
and to changes in parameters of G(s).
n
Benefits of feedback control
Despite the cost and increased system complexity, closed-loop feedback con-
trol has the following advantages:
se
• Decreased sensitivity of the system to variations in the parameters of the
process.
• Improved rejection of the disturbances.
as
• Improved measurement noise attenuation.
• Improved reduction of the steady-state error of the system.
• Easy control and adjustment of the transient response of the system.
H
Using the notion of a tracking error signal, it will be readily apparent that it
is possible to utilize feedback with a controller in the loop to improve system
performance.
as
E(s) = R(s) − Y (s).
Dr
2
With E(s) = R(s) − Y (s), we have
1 G(s) Gc (s)G(s)
E(s) = R(s) − Td (s) + N (s).
n
1 + Gc (s)G(s) 1 + Gc (s)G(s) 1 + Gc (s)G(s)
se
1 G(s) L(s)
E(s) = R(s) − Td (s) + N (s). (2)
1 + L(s) 1 + L(s) 1 + L(s)
We will see later on how L(s) plays a fundamental role in loop analysis.
as
Error signal analysis
We can now define the sensitivity and complementary sensitivity function
respectively in terms of L(s) as
1 L(s)
S(s) = and T (s) = .
Then,
1 + L(s)
H 1 + L(s)
and clearly, we cannot make both S(s) and T (s) simultaneously. Design com-
ye
1 G(s) L(s)
E(s) = R(s) − Td (s) + N (s).
1 + L(s) 1 + L(s) 1 + L(s)
3
3 Sensitivity
Sensitivity of Control Systems to Parameter Variations
n
• A process, represented by the transfer function G(s) is subject to a chang-
ing environment, aging, ignorance of the exact values of the process pa-
se
rameters, and other natural factors that affect a control process.
• In the open-loop system, all these errors and changes result in a changing
and inaccurate output.
• However, a closed-loop system senses the change in the output due to the
as
process changes and attempts to correct the output.
• The sensitivity of a control system to parameter variations is of prime im-
portance. A primary advantage of a closed-loop feedback control system
is its ability to reduce the system’s sensitivity.
H
Sensitivity of control systems to parameter variations
Consider (1) where Td (s) = N (s) = 0, and assume that Gc (s)G(s) >> 1 for
all frequencies of interest, then
Y (s) ≈ R(s).
d
Thus, increasing the loop gain (Gc (s)G(s)) reduces the effect of G(s) on the
output. Suppose the plant G(s) undergoes a change such that the true plant
model is G(s) + ∆G(s). The change in the plant may be due to a changing
ye
−Gc (s)∆G(s)
∆E(s) ≈ R(s).
(1 + L(s))2
Y (s)
T (s) = ,
R(s)
4
then the system sensitivity is defined as
∆T (s)/T (s)
S(s) = .
n
∆G(s)/G(s)
se
∂T /T ∂T G
S= = ·
∂G/G ∂G T
Sensitivity
as
For the open-loop system, the sensitivity to changes in the plant G(s) is 1.
For the closed-loop system, the transfer function is given by
Gc (s)G(s)
T (s) = .
1 + Gc (s)G(s)
or
T
SG =
∂T G
∂G T
· = H
Therefore the sensitivity of the closed-loop transfer function T (s) with respect
to G(s) is
Gc
·
G
(1 + Gc G)2 Gc G/(1 + Gc G)
d
T 1
SG =
1 + Gc (s)G(s)
Note that S above is exactly the same as the sensitivity function S(s) defined
ye
previously.
Sensitivity
Often, we seek to determine SαT , where α is a parameter within the transfer
function of a block G. Using the chain rule, we find that
Sa
SαT = SG
T G
Sα .
Another way of computing the sensitivity function for a system with transfer
function
N (s, α)
T (s, α) = ,
D(s, α)
where α is a parameter that is subject to variation, is through the following
Dr
formula:
SαT = SαN − SαD .
Thus by choosing a suitable controller Gc (s), we can construct a loop gain L(s)
and hence a closed-loop transfer function T (s) such that the system becomes
insensitive (or less sensitive) to variations in the plant transfer function.
5
n
se
Example
Calculate the sensitivity of the closed-loop transfer function to changes in
the parameter a. The closed loop transfer function is given by:
as
T (s) =
s2 + as + K
and the sensitivity
∂T a −as
SaT = · = 2
∂a T s + as + K
in a ?
4 Transient Response
H
How can we reduce the sensitivity of the closed-loop transfer function to changes
d
Transient Response
Transient Response
Let us consider this speed-control system: We previously determined the
transfer function relating the angular speed and the armature voltage as:
Dr
ω(s) K1
= G(s) = .
Va (s) τ1 s + 1
where
Km Ra J
K1 = and τ1 = .
Ra b + Kb Km Ra b + Kb Km
6
n
se
Transient Response
k2 E
If the open-loop motor is subjected to a step input Va (s) = s , the transient
as
speed change is given by:
• If this transient response is too slow, we must choose another motor with
H
a different time constant τ1 if possible.
• However, because τ1 is dominated by the load inertia J, it may not be
possible to achieve much alteration of the transient response.
Transient Response
d
Let us now consider the same scenario using a feedback loop. The closed-
ye
Ka K1
ω(t) = k2 E(1 − e−pt )
1 + Ka Kt K1
Dr
where p = (1 + Ka Kt K1 )/τ1 .
Transient Response
For a typical application, the open-loop pole may be 1/τ1 = 0.1, whereas the
closed-loop pole could be (Ka Kt K1 )/τ1 = 10. The response of the open-loop
and closed-loop system are compared below:
7
n
se
as
5 Steady-state error
Steady-state error
H
• The steady-state error is the error after the transient response has de-
cayed, leaving only the continuous response.
• The error of an open-loop system is given by:
1
Ec (s) = R(s)
1 + Gc (s)G(s)
s→0 s
Steady-state error
Dr
• The value of G(s) when s = 0 is often called the DC gain and is normally
greater than one in magnitude.
• Therefore, the open-loop system will usually have a steady-state error of
significant magnitude.
8
• By contrast, the closed-loop system with a reasonably large DC loop gain
L(0) = Gc(0)G(0) will have a small steady-state error.
n
• Why not “calibrate” the system such that the value of G(0) = 1 and then
there is no need for closed-loop feedback control ?
• What if the system parameters change due to parameter drift, environ-
se
mental factors ? We recalibrate ?
Steady-state error
Let us consider a unity feedback system with process transfer function:
as
K
G(s) = .
τs + 1
For a unit-step input, the steady-state error of the open-loop system is:
e(∞) = 1 − G(0) = 1 − K.
H
For the closed-loop system subjected to the same input:
e(∞) = 1 − T (0) = 1 −
K
1+K
=
1
1+K
.
where T (s) = G(s)/(1 + G(s)). For the open-loop system, we would choose
K = 1, while for the closed-loop system, we could choose K = 100.
d
What happens if there is a 10% change in parameter K ?
ye
6 Example
Example
A model of a boring machine control is shown where Y (s) is the actual
angle of direction of travel of the boring machine and R(s) is the desired angle.
The effect of load on the machine is represented by the disturbance Td (s). The
Sa
Dr
design objective is to select the gain K so that the response to input angle
changes is desirable while we maintain minimal error due to the disturbance.
The output is represented by:
K + 11s 1
Y (s) = R(s) + 2 Td (s).
s2 + 12s + K s + 12s + K
9
To reduce the effect of the disturbance, we would choose K to be large. Let’s
say we choose K = 100 and then we examine the response of the system. If we
apply a unit step input r(t) = 1, Td (t) = 0, we obtain
n
se
as
Example
H
Now, let us investigate the response of the system when a unit-step distur-
bance is applied, i.e., Td (t) = 1 and r(t) = 0. The effect of the disturbance is
d
ye
Sa
Example
We repeat the exercise before but this time by choosing a smaller value of
K = 20. What do we expect to see ?
Example
Dr
10
n
se
as
H
d
ye
1
Similarly, for Td (s) = s with R(s) = 0:
1 1 1
lim y(t) = lim sY (s) = s. . = .
t→∞ s→0 s2 + 12s + K s K
Sa
Example
Finally, let us examine the sensitivity of the system to a change in the process
G(s)
T 1 s(s + 1)
SG = = 2 .
1 + GH s + 12s + K
For low frequencies (|s| < 1),
Dr
T s
SG ≈ .
K
Given that K ≈ 20 − 100, it is clear that the closed-loop system is relatively
insensitive to changes in the process G(s) over a reasonable frequency range.
11