Lecture 5

Time-Domain Performance

1
 Outline

The Time-Domain Performance of Feedback Systems

 Test Input Signals

 Performance of Second-Order System

 Effects of a Third Pole and a Zero on the Second-Order System Response

 The s-Plane Root Location and the Transient Response

 The Steady-State Error of Feedback Control Systems

 System Simulation Using Matlab

2
 Overview
 Easy control and adjustment of the transient and steady-state response of a control system is a distinct
advantage of feedback control systems;
 To analyze and design a control system, we must define and measure its performance, the controller
parameters may be adjusted to provide the desired response which is often described by design
specifications.
 Control systems are inherently dynamic, their performance is usually specified in terms of both the transient
response the steady-state response.
- Transient response is the response that disappears with time;
- Steady-state response is the response that exists for a long time following an input signal initiation.

Closed-loop System

𝒀𝒀(𝒔𝒔) 𝑮𝑮𝒄𝒄 𝑮𝑮
𝑻𝑻 𝒔𝒔 = =
𝑹𝑹(𝒔𝒔) 𝟏𝟏 + 𝑮𝑮𝒄𝒄 𝑮𝑮𝑮𝑮

3
 Test Input Signals
Control systems are inherently time-domain systems, so the system transient or time
performance is the response of prime interest for control systems.

 Is the system stable? (will be discussed in the following lectures)

 If stable, how to measure and compare the performance of several competing designs?
- Provide several measures of performance (response time, percent overshoot etc.)
- Test the system by standard test input signals.

Three Standard Test Input Signals

4
 Test Input Signal in Time and s-Domain
 There is a reasonable correlation between the response of a system to a standard test input and the
system’s ability to perform under normal operating conditions.

 Many control systems experience input signals that are very similar to the standard test signals.

Step Ramp Parabolic

Impulse

𝒓𝒓 𝒕𝒕 = 𝒕𝒕𝒏𝒏
𝒏𝒏!
𝑹𝑹 𝒔𝒔 =
𝒔𝒔𝒏𝒏+𝟏𝟏
except for impulse input:
𝑟𝑟 𝑡𝑡 = 𝛿𝛿 𝑡𝑡 , 𝑅𝑅 𝑠𝑠 = 1
5
 Second-Order System
𝐺𝐺 𝑠𝑠
𝑌𝑌 𝑠𝑠 = 𝑅𝑅(𝑠𝑠)
1 + 𝐺𝐺 𝑠𝑠

𝝎𝝎𝒏𝒏 𝟐𝟐 𝜔𝜔𝑛𝑛 : Natural Frequency;

𝒀𝒀 𝒔𝒔 = 𝟐𝟐 𝑹𝑹(𝒔𝒔) 𝜁𝜁 : Damping Ratio.
𝒔𝒔 + 𝟐𝟐𝜻𝜻𝝎𝝎𝒏𝒏 𝒔𝒔 + 𝝎𝝎𝒏𝒏 𝟐𝟐

6
 Time Response to Impulse Input

𝑅𝑅 𝑠𝑠 = 1

𝜔𝜔𝑛𝑛 2
𝑌𝑌 𝑠𝑠 = 2
𝑠𝑠 + 2𝜁𝜁𝜔𝜔𝑛𝑛 𝑠𝑠 + 𝜔𝜔𝑛𝑛 2

𝝎𝝎𝒏𝒏 −𝜻𝜻𝝎𝝎 𝒕𝒕
𝒚𝒚 𝒕𝒕 = 𝒆𝒆 𝒏𝒏 𝐬𝐬𝐬𝐬𝐬𝐬(𝝎𝝎𝒏𝒏 𝜷𝜷𝜷𝜷)
𝜷𝜷

where 𝛽𝛽 = 1 − 𝜁𝜁 2 ,

0 < 𝜁𝜁 < 1.

7
 Time Response to Step Input

1
𝑅𝑅 𝑠𝑠 =
𝑠𝑠
𝜔𝜔𝑛𝑛 2 1
𝑌𝑌 𝑠𝑠 = 2
𝑠𝑠 + 2𝜁𝜁𝜔𝜔𝑛𝑛 𝑠𝑠 + 𝜔𝜔𝑛𝑛 2 𝑠𝑠

𝟏𝟏 −𝜻𝜻𝝎𝝎 𝒕𝒕
𝒚𝒚 𝒕𝒕 = 𝟏𝟏 − 𝒆𝒆 𝒏𝒏 𝐬𝐬𝐬𝐬𝐬𝐬(𝝎𝝎𝒏𝒏 𝜷𝜷𝜷𝜷 + 𝜽𝜽)
𝜷𝜷

where 𝛽𝛽 = 1 − 𝜁𝜁 2 ,

𝜃𝜃 = cos −1 𝜁𝜁

0 < 𝜁𝜁 < 1.

8
 Standard Performance Measures
Standard performance measures are often defined in terms of the unit step response of the
closed-loop system.
Peak Value: 𝑀𝑀𝑝𝑝𝑝𝑝 Peak Time:
𝜋𝜋
Final Value: 𝑓𝑓𝑣𝑣 =1 𝑇𝑇𝑝𝑝 =
1 𝜔𝜔𝑛𝑛 1 − 𝜁𝜁 2
Time Constant: 𝜏𝜏 =
𝜁𝜁𝜔𝜔𝑛𝑛 Rise Time:
2.16𝜁𝜁 + 0.60
𝑇𝑇𝑟𝑟𝑟 =
𝜔𝜔𝑛𝑛
(0.3 < 𝜁𝜁 < 0.8)

2% Settling Time:
4
𝑇𝑇𝑠𝑠 ≅
𝜁𝜁𝜔𝜔𝑛𝑛

Percent Overshoot:
−𝜁𝜁𝜁𝜁
�
𝑃𝑃. 𝑂𝑂. = 100𝑒𝑒 1−𝜁𝜁 2

𝑀𝑀𝑝𝑝𝑝𝑝−𝑓𝑓𝑓𝑓
𝑃𝑃. 𝑂𝑂. = × 100%
𝑓𝑓𝑓𝑓

9
P.O. and Normalized Peak Time vs. 𝜻𝜻

Need
Compromise!

10
Effects of 𝝎𝝎𝒏𝒏 and 𝜻𝜻 on The Step Response
with 𝜁𝜁 =0.2, different 𝜔𝜔𝑛𝑛 with 𝜔𝜔𝑛𝑛 = 5, different 𝜁𝜁

11
 Example
Consider the following system, select gain K and the parameter p so that the time-domain specifications to a
unit step input are satisfied.
 Specifications: 2% settling time 𝑇𝑇𝑠𝑠 ≤ 4 𝑠𝑠; and percent overshoot 𝑃𝑃. 𝑂𝑂. ≤ 5%.

Step 1. Transfer function:

𝐾𝐾 𝜔𝜔𝑛𝑛 2
𝑇𝑇 𝑠𝑠 = 2 (= 2 )
𝑠𝑠 + 𝑝𝑝𝑝𝑝 + 𝐾𝐾 𝑠𝑠 + 2𝜁𝜁𝜔𝜔𝑛𝑛 𝑠𝑠 + 𝜔𝜔𝑛𝑛 2
2𝜁𝜁𝜁𝜁𝑛𝑛 = 𝑝𝑝, 𝜔𝜔𝑛𝑛 2 = 𝐾𝐾
Step 2. To satisfy settling time requirement:
4
≤4 𝜁𝜁𝜁𝜁𝑛𝑛 ≥ 1
𝜁𝜁𝜁𝜁𝑛𝑛
12
Step 3. To satisfy the P.O. requirement:

𝑃𝑃. 𝑂𝑂. ≤ 5% 𝜁𝜁 ≥ 0.69

Step 4. Choose suitable values:

Can choose 𝜁𝜁𝜁𝜁𝑛𝑛 = 1 𝑝𝑝 = 2

𝜔𝜔𝑛𝑛 = 2
1
𝜁𝜁 = 0.707 = 𝐾𝐾 = 2
2
13
Specifications and Root Locations

𝐾𝐾
𝑇𝑇 𝑠𝑠 = 2
𝑠𝑠 + 2𝑠𝑠 + 2
Poles: −1 ± 𝑗𝑗𝑗

𝜔𝜔𝑛𝑛 2
𝑇𝑇 𝑠𝑠 = 2
𝑠𝑠 + 2𝜁𝜁𝜔𝜔𝑛𝑛 𝑠𝑠 + 𝜔𝜔𝑛𝑛 2

Poles:
𝑝𝑝1,2 = −𝜁𝜁𝜔𝜔𝑛𝑛 ± 𝑗𝑗𝜔𝜔𝑛𝑛 1 − 𝜁𝜁 2

14
 Effects of A Third Pole
Assume 𝜔𝜔𝑛𝑛 = 1, Consider a system with two complex poles and an additional pole

1
𝑇𝑇 𝑠𝑠 = 2
(𝑠𝑠 +2𝜁𝜁𝜔𝜔𝑛𝑛 𝑠𝑠 + 1)(𝛾𝛾𝛾𝛾 + 1)

The time response of a third-order system can be approximated by the dominant roots of the second-order
system as long as the real part of the dominant roots is less than one tenth of the real part of the third pole.

𝟏𝟏� ≥ 𝟏𝟏𝟏𝟏 𝜻𝜻𝝎𝝎

𝜸𝜸 𝒏𝒏

NOTE: DC gain 𝑇𝑇(0) should be kept the

same after approximation.

15
 Example
System 1.
System 2.

System 1 can be
System 3. approximated by System 2,
while System 3 can NOT!

16
 Effects of A Finite Zero

Consider a system:

𝜔𝜔𝑛𝑛 2
(𝑠𝑠 + 𝑎𝑎)
𝑇𝑇 𝑠𝑠 = 2 𝑎𝑎
𝑠𝑠 + 2𝜁𝜁𝜔𝜔𝑛𝑛 𝑠𝑠 + 𝜔𝜔𝑛𝑛 2

If 𝑎𝑎 ≫ 𝜁𝜁𝜔𝜔𝑛𝑛 :

the system can be simplified

as:
𝜔𝜔𝑛𝑛 2
𝑇𝑇 𝑠𝑠 ≈ 2
𝑠𝑠 + 2𝜁𝜁𝜔𝜔𝑛𝑛 𝑠𝑠 + 𝜔𝜔𝑛𝑛 2

17
 Example

1.6(𝑠𝑠 + 2.5) 𝜔𝜔𝑛𝑛 2

(𝑠𝑠 + 𝑎𝑎)
𝑇𝑇 𝑠𝑠 = 2 𝑇𝑇 𝑠𝑠 = 2 𝑎𝑎
(𝑠𝑠 +6𝑠𝑠 + 25)(𝑠𝑠 + 0.16) (𝑠𝑠 +2𝜁𝜁𝜔𝜔𝑛𝑛 𝑠𝑠 + 𝜔𝜔𝑛𝑛 2 )(1 + 𝜏𝜏𝜏𝜏)

𝜁𝜁𝜔𝜔𝑛𝑛 = 3, 𝑎𝑎 = 2.5, 𝜏𝜏 = 1/0.16

For this system, zero and third pole can NOT be

neglected!

For the actual third-order system:

Ts=1.6s, P.O.=38%

𝟐𝟐𝟐𝟐
For the second-order 𝑻𝑻 𝒔𝒔 =
𝒔𝒔𝟐𝟐 + 𝟔𝟔𝟔𝟔 + 𝟐𝟐𝟐𝟐
system:
Ts=1.33s, P.O.=9.5%
18
 Quiz

Consider the following system, can we neglect the effects of third

pole? If yes, obtain approximated transfer function and estimate P.O.

19
 Quiz

For a second order system, determine the root locations

in s-plane which satisfies:

1. 10% < P.O. < 20%

2. Settling time < 0.6.

20
 The s-Plane Location and The Transient Response
Transfer function for a closed-loop system can be written as:
𝑌𝑌(𝑠𝑠) ∑ 𝑃𝑃𝑖𝑖 (𝑠𝑠)∆𝑖𝑖 (𝑠𝑠)
𝑇𝑇 𝑠𝑠 = =
𝑅𝑅(𝑠𝑠) ∆(𝑠𝑠)

Characteristic equation of the system: ∆ 𝑠𝑠 = 0

For a unit feedback control system: ∆ 𝒔𝒔 = 𝟏𝟏 + 𝑮𝑮𝒄𝒄 𝒔𝒔 𝑮𝑮 𝒔𝒔 = 𝟎𝟎

Time response of a system depends on the poles and zeros of its transfer function 𝑇𝑇 𝑠𝑠 ; while
for a closed-loop system, the poles of are the roots of the characteristic equation: ∆ 𝑠𝑠 .

Transfer function:

𝑌𝑌(𝑠𝑠) 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺(𝑠𝑠)

𝑇𝑇 𝑠𝑠 = =
𝑅𝑅(𝑠𝑠) 1 + 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠 𝐻𝐻(𝑠𝑠)

21
 Time Response of System: General Form
If the system (with DC gain = 1) has no repeated roots, its unit step response can be
formulated as a partial fraction expansion as:
𝑀𝑀 𝑁𝑁
1 𝐴𝐴𝑖𝑖 𝐵𝐵𝑘𝑘 𝑠𝑠 + 𝐶𝐶𝑘𝑘
𝑌𝑌 𝑠𝑠 = +� +� 2
𝑠𝑠 𝑠𝑠 + 𝜎𝜎𝑖𝑖 𝑠𝑠 + 2𝛼𝛼𝑘𝑘 𝑠𝑠 + (𝛼𝛼𝑘𝑘 2 + 𝜔𝜔𝑘𝑘 2 )
𝑖𝑖=1 𝐾𝐾=1

where 𝐴𝐴𝑖𝑖 , 𝐵𝐵𝑘𝑘 and 𝐶𝐶𝑘𝑘 are constants; the roots of the system must be either
𝑠𝑠 = −𝜎𝜎𝑖𝑖 or 𝑠𝑠 = −𝛼𝛼𝑘𝑘 ± 𝑗𝑗𝜔𝜔𝑘𝑘
The transient response can be obtained by inverse Laplace transform:
𝑀𝑀 𝑁𝑁

𝑦𝑦 𝑡𝑡 = 1 + � 𝐴𝐴𝑖𝑖 𝑒𝑒 −𝜎𝜎𝑖𝑖𝑡𝑡 + � 𝐷𝐷𝑘𝑘 𝑒𝑒 −𝛼𝛼𝑘𝑘 𝑡𝑡 sin(𝜔𝜔𝑘𝑘 𝑡𝑡 + 𝜃𝜃𝑘𝑘 )

𝑖𝑖=1 𝑘𝑘=1

exponential
Steady-state terms Damped sinusoidal
output terms
where 𝐷𝐷𝑘𝑘 is a constant depends on 𝐵𝐵𝑘𝑘 , 𝐶𝐶𝑘𝑘 , 𝛼𝛼𝑘𝑘 and 𝜔𝜔𝑘𝑘 .

For the response to be stable (bounded for a step input) – the real part of the poles
must be in the left-hand portion of the s-plane.
22
Step Response for Various Root Locations in the s-Plane

𝑴𝑴

𝒚𝒚 𝒕𝒕 = 𝟏𝟏 + � 𝑨𝑨𝒊𝒊 𝒆𝒆−𝝈𝝈𝒊𝒊 𝒕𝒕
𝒊𝒊=𝟏𝟏
𝑵𝑵

+ � 𝑫𝑫𝒌𝒌 𝒆𝒆−𝜶𝜶𝒌𝒌 𝒕𝒕 𝐬𝐬𝐬𝐬𝐬𝐬(𝝎𝝎𝒌𝒌 𝒕𝒕 + 𝜽𝜽𝒌𝒌 )

𝒌𝒌=𝟏𝟏

(𝑠𝑠 = −𝜎𝜎𝑖𝑖 ) (𝑠𝑠 = −𝛼𝛼𝑘𝑘 ± 𝑗𝑗𝜔𝜔𝑘𝑘 )

(The conjugate root is not shown in

this figure.)

23
 Root Location and System Design

 It is important for the control system designer to understand the complete relationship of the
frequency domain representation of a linear system, the poles and zeros of its transfer
function, and its time-domain response to step and other inputs;

 In such areas as signal processing and control, many analysis and design calculations are
done in the s-plane, where a system model is represented in terms of the poles and zeros of
its transfer function;

 The control system designer will envision the effects of the step and impulse response of
adding, deleting, or moving poles and zeros of 𝑇𝑇(𝑠𝑠) in the s-plane;

 An experienced designer is aware of the effects of zero locations on system response. For
example, moving a zero closer to a specific pole will reduce the relative contribution to the
output response. In other words, if there is a zero near the pole at 𝑠𝑠 = −𝜎𝜎𝑖𝑖 , then 𝐴𝐴𝑖𝑖 will be
much smaller in magnitude.

24
The Steady-State Error of Feedback Control System

One of the fundamental reasons for using feedback, despite its cost and increased
complexity, is the attendant improvement in the reduction of the steady-state error of the
system. Consider a unit negative feedback system (𝐻𝐻(𝑠𝑠) = 1), in the absence of external
disturbances (𝑇𝑇𝑑𝑑 𝑠𝑠 = 0) and measurement noise (𝑁𝑁 𝑠𝑠 = 0), tracking error is:

1
𝐸𝐸 𝑠𝑠 = 𝑅𝑅 𝑠𝑠 − 𝑌𝑌 𝑠𝑠 = 𝑅𝑅(𝑠𝑠)
1 + 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠

Using the final value theorem, the steady-state error is:

1
lim 𝑒𝑒 𝑡𝑡 = 𝑒𝑒𝑠𝑠𝑠𝑠 = lim 𝑠𝑠 𝑅𝑅(𝑠𝑠)
𝑡𝑡→∞ 𝑠𝑠→0 1 + 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠

25
 Steady-State Error to Step Inputs

 Step Input of magnitude 𝐴𝐴:

𝐴𝐴� 𝐴𝐴
𝑠𝑠
𝑒𝑒𝑠𝑠𝑠𝑠 = lim 𝑠𝑠 =
𝑠𝑠→0 1 + 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠 1 + lim 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠
𝑠𝑠→0

The loop transfer function can be written in general form as

𝐾𝐾 ∏𝑀𝑀
𝑖𝑖=1(𝑠𝑠 + 𝑧𝑧𝑖𝑖 )
𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠 = 𝑄𝑄 where 𝑧𝑧𝑖𝑖 ≠ 0, 𝑝𝑝𝑘𝑘 ≠ 0.
𝑠𝑠 𝑵𝑵 ∏𝑘𝑘=1(𝑠𝑠 + 𝑝𝑝𝑘𝑘 )

26
 Steady-State Error to Step Inputs
The number of integration indicates a system with type number that is equal to
𝑵𝑵, which determines the steady-state error of the system.

Given the position error constant: 𝐾𝐾𝑝𝑝 = lim 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠

𝑠𝑠→0

𝐴𝐴
 For a type-zero system (𝑁𝑁 = 0): 𝑒𝑒𝑠𝑠𝑠𝑠 =
1 + 𝐾𝐾𝑝𝑝

𝐴𝐴 𝐴𝐴𝑠𝑠 𝑁𝑁
 For a type-N system with 𝑁𝑁 ≥ 1: 𝑒𝑒𝑠𝑠𝑠𝑠 = lim = lim = 0.
𝑠𝑠→0 ∏ 𝑧𝑧𝑖𝑖 𝑠𝑠→0 𝑁𝑁 ∏ 𝑧𝑧𝑖𝑖
1 + 𝐾𝐾 𝑁𝑁 𝑠𝑠 + 𝐾𝐾
𝑠𝑠 ∏ 𝑝𝑝𝑘𝑘 ∏ 𝑝𝑝𝑘𝑘

27
 Steady-State Error to Ramp Inputs

 Ramp Input with a slope 𝐴𝐴:

𝐴𝐴�
𝑠𝑠 2 𝐴𝐴 𝐴𝐴
𝑒𝑒𝑠𝑠𝑠𝑠 = lim 𝑠𝑠 = =
𝑠𝑠→0 1 + 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠 𝑠𝑠 + lim 𝑠𝑠𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠 lim 𝑠𝑠𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠
𝑠𝑠→0 𝑠𝑠→0

𝐾𝐾 ∏𝑀𝑀
𝑖𝑖=1(𝑠𝑠 + 𝑧𝑧𝑖𝑖 )
𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠 = 𝑄𝑄 where 𝑧𝑧𝑖𝑖 ≠ 0, 𝑝𝑝𝑘𝑘 ≠ 0.
𝑠𝑠 𝑵𝑵 ∏𝑘𝑘=1(𝑠𝑠 + 𝑝𝑝𝑘𝑘 )

28
 Steady-State Error to Ramp Inputs

Denote the velocity error constant: 𝐾𝐾𝑣𝑣 = lim 𝑠𝑠𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠

𝑠𝑠→0

 For a type-zero system (𝑁𝑁 = 0): 𝑒𝑒𝑠𝑠𝑠𝑠 = ∞

𝐴𝐴 𝐴𝐴 𝐴𝐴
 For a type-one system (𝑁𝑁 = 1): 𝑒𝑒𝑠𝑠𝑠𝑠 = lim = =
𝑠𝑠→0 ∏(𝑠𝑠 + 𝑧𝑧𝑖𝑖 ) ∏ 𝑧𝑧𝑖𝑖 𝐾𝐾𝑣𝑣
𝑠𝑠𝑠𝑠 𝐾𝐾
𝑠𝑠 ∏(𝑠𝑠 + 𝑝𝑝𝑘𝑘 ) ∏ 𝑝𝑝𝑘𝑘

𝐴𝐴 𝐴𝐴𝑠𝑠 𝑵𝑵−𝟏𝟏
 For a type-N system with 𝑁𝑁 > 1: 𝑒𝑒𝑠𝑠𝑠𝑠 = lim = =0
𝑠𝑠→0 ∏(𝑠𝑠 + 𝑧𝑧𝑖𝑖 ) ∏ 𝑧𝑧𝑖𝑖
𝑠𝑠𝑠𝑠 𝑵𝑵 𝐾𝐾
𝑠𝑠 ∏ (𝑠𝑠 + 𝑝𝑝𝑘𝑘 ) ∏ 𝑝𝑝𝑘𝑘

29
 Steady-State Error to Acceleration Inputs

 Acceleration Input 𝑅𝑅 𝑠𝑠 = 𝐴𝐴/𝑠𝑠 3 (𝑟𝑟 𝑡𝑡 = 𝐴𝐴𝑡𝑡 2 /2):

𝐴𝐴�
𝑠𝑠 3 𝐴𝐴 𝐴𝐴
𝑒𝑒𝑠𝑠𝑠𝑠 = lim 𝑠𝑠 = 2 =
𝑠𝑠→0 1 + 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠 𝑠𝑠 + lim 𝑠𝑠 2 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠 lim 𝑠𝑠 2 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠
𝑠𝑠→0 𝑠𝑠→0

𝐾𝐾 ∏𝑀𝑀
𝑖𝑖=1(𝑠𝑠 + 𝑧𝑧𝑖𝑖 )
𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠 = 𝑄𝑄 where 𝑧𝑧𝑖𝑖 ≠ 0, 𝑝𝑝𝑘𝑘 ≠ 0.
𝑠𝑠 𝑵𝑵 ∏𝑘𝑘=1(𝑠𝑠 + 𝑝𝑝𝑘𝑘 )

30
 Steady-State Error to Acceleration Inputs

Denote the acceleration error constant: 𝐾𝐾𝑎𝑎 = lim 𝑠𝑠 2 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠

𝑠𝑠→0

 For a type-N system with 𝑁𝑁 < 2: 𝑒𝑒𝑠𝑠𝑠𝑠 = ∞

𝐴𝐴 𝐴𝐴 𝐴𝐴
𝑒𝑒𝑠𝑠𝑠𝑠 = lim = =
 For a type-two system (𝑁𝑁 = 2): 𝑠𝑠→0 2 ∏(𝑠𝑠 + 𝑧𝑧𝑖𝑖 ) ∏ 𝑧𝑧𝑖𝑖 𝐾𝐾𝑎𝑎
𝑠𝑠 𝐾𝐾 2 𝐾𝐾
𝑠𝑠 ∏(𝑠𝑠 + 𝑝𝑝𝑘𝑘 ) ∏ 𝑝𝑝𝑘𝑘

 For a type-N system with 𝑁𝑁 > 2: 𝐴𝐴 𝐴𝐴𝑠𝑠 𝑵𝑵−𝟐𝟐

𝑒𝑒𝑠𝑠𝑠𝑠 = lim = =0
𝑠𝑠→0 2 ∏(𝑠𝑠 + 𝑧𝑧𝑖𝑖 ) ∏ 𝑧𝑧𝑖𝑖
𝑠𝑠 𝐾𝐾 𝑵𝑵 𝐾𝐾
𝑠𝑠 ∏(𝑠𝑠 + 𝑝𝑝𝑘𝑘 ) ∏ 𝑝𝑝𝑘𝑘

31
 Summary Table

 The control system error constants 𝑲𝑲𝒑𝒑 , 𝑲𝑲𝒗𝒗 and 𝑲𝑲𝒂𝒂 , describe the ability of a system to reduce or eliminate the
steady-state error. Therefore, they are utilized as numerical measure of the steady-state performance. The designer
determines the error constants for a given system and attempts to determine methods of increasing the error
constants while maintaining an acceptable transient response.
32
 Example – Mobile Robot Steering Control

Consider the following system of mobile robot. Transfer function of controller is

𝐺𝐺𝑐𝑐 𝑠𝑠 = 𝐾𝐾1 + 𝐾𝐾2 /𝑠𝑠

𝐾𝐾(𝐾𝐾1 𝑠𝑠 + 𝐾𝐾2 )
Loop transfer function: 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺(𝑠𝑠) =
𝜏𝜏𝑠𝑠 2 + 𝑠𝑠

33
 Example – Mobile Robot Steering Control

𝐾𝐾𝐾𝐾
 When 𝐾𝐾2 = 0 -> 1
𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺(𝑠𝑠) = 𝜏𝜏𝜏𝜏+1 -> type-0 system:

For step input: where 𝐾𝐾𝑝𝑝 = 𝑠𝑠→0

lim 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺 𝑠𝑠 = 𝐾𝐾𝐾𝐾1
For ramp input: 𝑒𝑒𝑠𝑠𝑠𝑠 = ∞

𝐾𝐾(𝐾𝐾1 𝑠𝑠+𝐾𝐾2 )
 When 𝐾𝐾2 > 0 -> 𝐺𝐺𝑐𝑐 𝑠𝑠 𝐺𝐺(𝑠𝑠) = -> type-1 system:
𝑠𝑠(𝜏𝜏𝜏𝜏+1)
For step input: 𝑒𝑒𝑠𝑠𝑠𝑠 = 0

For ramp input: where

34
Transient response of the system to a triangular wave input when 𝐾𝐾2 > 0

35
Example – Steady-State Error for A Nonunity Negative Feedback System

Consider the following system, determine K so that the ESS for a unit step input is minimized.

36
Example – Steady-State Error for A Nonunity Negative Feedback System

Solutions:
𝐸𝐸 𝑠𝑠 = 𝑅𝑅 𝑠𝑠 − 𝑌𝑌 𝑠𝑠 = [1 − 𝑇𝑇(𝑠𝑠)]𝑅𝑅(𝑠𝑠)

For a unit step input: 𝑒𝑒𝑠𝑠𝑠𝑠 = lim 𝑠𝑠𝑠𝑠 𝑠𝑠 = 1 − 𝑇𝑇(0)

𝑠𝑠→0

To minimize ESS, it requires:

Therefore:

37
 Performance Index
A performance index is a quantitative measure of the performance of a system and is
chosen so that emphasis is given to the important system specifications.
A system is considered an optimum control system when the system parameters are adjusted so
that the index reaches an extremum, commonly a minimum value.

Common performance index:

General form:

38
39
Optimum Coefficients for T(s) based on ITAE Criterion

40
 Design Example – Hubble Space Telescope Control
For the following control system, choose 𝑲𝑲𝟏𝟏 and 𝑲𝑲, to satisfy:
(1) Percent overshoot of the output to a step command r(t) is 𝑃𝑃. 𝑂𝑂. ≤ 10%
(2) Steady-state error to a ramp command is minimized;
(3) Effect of a step disturbance is reduced.

41
Design Example – Hubble Space Telescope Control

Step 1. Re-arrange the block diagram to achieve a standard form.

42
Step 2. Obtain Y(s) and E(s) in terms of R(s), Td(s) and system parameters.

𝐾𝐾𝐾𝐾(𝑠𝑠) 𝐺𝐺(𝑠𝑠)
𝑌𝑌 𝑠𝑠 = 𝑅𝑅 𝑠𝑠 + 𝑇𝑇 𝑠𝑠
1 + 𝐾𝐾𝐾𝐾(𝑠𝑠) 1 + 𝐾𝐾𝐾𝐾 𝑠𝑠 𝑑𝑑

1 𝐺𝐺(𝑠𝑠)
𝐸𝐸 𝑠𝑠 = 𝑅𝑅 𝑠𝑠 − 𝑌𝑌 𝑠𝑠 = 𝑅𝑅 𝑠𝑠 − 𝑇𝑇 𝑠𝑠
1 + 𝐾𝐾𝐾𝐾(𝑠𝑠) 1 + 𝐾𝐾𝐾𝐾(𝑠𝑠) 𝑑𝑑

43
Step 3. Consider requirement (1): 𝑃𝑃. 𝑂𝑂. ≤ 10% for a step input.

Characteristic equation of the system is

1 + 𝐾𝐾𝐾𝐾 𝑠𝑠 = 0 𝑠𝑠 2 + 𝐾𝐾1 𝑠𝑠 + 𝐾𝐾 = 0

2𝜁𝜁𝜔𝜔𝑛𝑛 = 𝐾𝐾1 , 𝜔𝜔𝑛𝑛 2 = 𝐾𝐾. Standard form:𝑠𝑠 2 + 2𝜁𝜁𝜔𝜔𝑛𝑛 𝑠𝑠 + 𝜔𝜔𝑛𝑛 2

For 𝑃𝑃. 𝑂𝑂. ≤ 10%, it must be satisfied that 𝜁𝜁 ≥ 0.6. we choose 𝜁𝜁=0.6, therefore

−𝜁𝜁𝜁𝜁 𝐾𝐾1
� = 𝐾𝐾
𝑃𝑃. 𝑂𝑂. = 100𝑒𝑒 1−𝜁𝜁 2 1.2

44
 1 𝐺𝐺(𝑠𝑠)
𝐸𝐸 𝑠𝑠 = 𝑅𝑅 𝑠𝑠 − 𝑌𝑌 𝑠𝑠 = 𝑅𝑅 𝑠𝑠 − 𝑇𝑇 𝑠𝑠
1 + 𝐾𝐾𝐾𝐾(𝑠𝑠) 1 + 𝐾𝐾𝐾𝐾(𝑠𝑠) 𝑑𝑑

Step 4. Consider requirement (2): minimize ESS to a ramp input: assume 𝑇𝑇𝑑𝑑 𝑠𝑠 = 0

1 𝑠𝑠 2 + 𝐾𝐾1 𝑠𝑠 𝐴𝐴
𝐸𝐸 𝑠𝑠 = 𝑅𝑅 𝑠𝑠 = 2
1 + 𝐾𝐾𝐾𝐾(𝑠𝑠) 𝑠𝑠 + 𝐾𝐾1 𝑠𝑠 + 𝐾𝐾 𝑠𝑠 2
𝐴𝐴
𝑒𝑒𝑠𝑠𝑠𝑠 = lim 𝑠𝑠𝑠𝑠 𝑠𝑠 =
𝑠𝑠→0 𝐾𝐾/𝐾𝐾1

To minimize ESS, we need large value of 𝐾𝐾/𝐾𝐾1 .

Step 5. Consider requirement (3): minimize ESS to a step disturbance: assume 𝑅𝑅 𝑠𝑠 = 0

𝐺𝐺 𝑠𝑠 1 𝐵𝐵
𝐸𝐸 𝑠𝑠 = − 𝑇𝑇𝑑𝑑 𝑠𝑠 = − 2
1 + 𝐾𝐾𝐾𝐾 𝑠𝑠 𝑠𝑠 + 𝐾𝐾1 𝑠𝑠 + 𝐾𝐾 𝑠𝑠
𝐵𝐵
𝑒𝑒𝑠𝑠𝑠𝑠 = lim 𝑠𝑠𝑠𝑠 𝑠𝑠 = −
𝑠𝑠→0 𝐾𝐾

To minimize ESS, we need large value of 𝐾𝐾.

45
Step 6. Choose suitable values.

𝐾𝐾1 To minimize ESS, we need large value of 𝐾𝐾/𝐾𝐾1 .

= 𝐾𝐾;
1.2 To minimize ESS, we need large value of 𝐾𝐾.
𝐾𝐾
Can choose 𝐾𝐾 = 100, then according to 1.21 = 𝐾𝐾, 𝐾𝐾1 = 12, and 𝐾𝐾/𝐾𝐾1 = 8.33.

Therefore,
𝐴𝐴
ESS for a ramp input is ≈ 0.12𝐴𝐴, ESS
8.33
for a step disturbance is
𝐵𝐵
− = −0.01𝐵𝐵.
100
All the requirements have been satisfied.

46
47