153
1 The Principle of Automat The Principle of Automat ic Control ic Control 自自自自自自 自自自自自自 Lecturers Lecturers Prof. Jiang B Prof. Jiang B in in Dr. Lu Dr. Lu Ningyun Ningyun College of Automation En College of Automation En gineering gineering NUAA NUAA 2008. Autumn 2008. Autumn

The Principle of Automatic Control 自动控制原理

  • Upload
    waylon

  • View
    347

  • Download
    2

Embed Size (px)

DESCRIPTION

The Principle of Automatic Control 自动控制原理. Lecturers : Prof. Jiang Bin Dr. Lu Ningyun College of Automation Engineering NUAA , 2008. Autumn. NUAA- The Principle of Automatic Control. Chapter 3 Time-domain analysis of control systems 控制系统时域分析. Review of Chap2. - PowerPoint PPT Presentation

Citation preview

Page 1: The Principle of Automatic Control 自动控制原理

11

The Principle of Automatic CoThe Principle of Automatic Controlntrol

自动控制原理自动控制原理LecturersLecturers :: Prof. Jiang BinProf. Jiang Bin

Dr. Lu NingyunDr. Lu NingyunCollege of Automation EngineeCollege of Automation Enginee

ringringNUAANUAA ,, 2008. Autumn2008. Autumn

Page 2: The Principle of Automatic Control 自动控制原理

2

NUAA-The Principle of Automatic Control

Chapter 3

Time-domain analysis of control systems

控制系统时域分析

Page 3: The Principle of Automatic Control 自动控制原理

3

The Principle of Automatic Control 2008

Review of Chap2• Modeling of control systems

– Differential equation model• How to obtain differential equation model?• Linearization a nonlinear model• Solving differential model is difficult

– Laplace and inverse Laplace transform• Formula• Properties• How to use laplace transform to solve differential model

– Transfer function

Page 4: The Principle of Automatic Control 自动控制原理

4

The Principle of Automatic Control 2008

Review of Chap2

– Transfer function• What is Transfer function?• What types of system can be characterized

by transfer functions?• How to find transfer function?• polynomial form and zero-pole-gain form• Relationship between poles and responses

– Modeling a complex control system by structure diagram or signal-flow graph

Page 5: The Principle of Automatic Control 自动控制原理

5

The Principle of Automatic Control 2008

Chap 3 Time-domain analysis of control systems

3-1 Time-domain criteria ( 时域指标) of control systems

3-2 Time response of a first-order system (一阶系统)

3-3 Time response of a prototype 2nd-order system ( 2 阶原型系统)

3-4 Analysis of higher-order system3-5 Stability of linear control systems3-6 Steady-state error (稳态误差)

Page 6: The Principle of Automatic Control 自动控制原理

6

The Principle of Automatic Control 2008

3-1 Time-domain criteria of control systems

Page 7: The Principle of Automatic Control 自动控制原理

7

The Principle of Automatic Control 2008

What is time-response( 时间响应) ? Since time is used as an independent variable

in most control system, it is usually of interest to evaluate the output response with respect to time, or simply, the time response.

• Typical input signals:– 单位阶跃 1(t)– 单位斜坡 t– 单位加速度 0.5t^2– 单位脉冲 δ(t)– 正弦函数 Asin(wt)

为了便于分析和设计;以及各种控制系统的性能比较,需要基本的输入函

数形式

Page 8: The Principle of Automatic Control 自动控制原理

8

The Principle of Automatic Control 2008

t0

1

Output c(t)

Time response of a control system is usually divided into transient process and steady-state process

Steady-state process (稳态过程) the part of the time response that remains after the transient has died out, 提供了稳态误差(steady-state error) 信息

Transient process ( 瞬态过程) ,transition from initial condition to the final condition由系统特性,瞬态过程可能衰减(decaying) 、发散 (divergence)( 或振荡 (oscillation) 等形式--提供系统稳定性 (stability) 的信息瞬态过程同时提供响应速度 (speed) 、阻尼 (damping) 等信息

Transient process and steady-state process

Page 9: The Principle of Automatic Control 自动控制原理

9

The Principle of Automatic Control 2008

Time-domain criteria

• 稳定性 (stability)

• 动态性能指标 (transient performance)– 响应速度 (speed of response)

• Rise time( 上升时间 ) , peak time( 峰值时间 ) , settling time( 调节时间 )

– 响应形式 (type of response)• Damping ( 阻尼 ) ratio, overshoot( 超调量 ) , period of osc

illation( 振荡周期)

• 稳态性能指标 (steady state performance) – 稳态误差,反映系统控制精度或抗扰动能力 (disturba

nce resistance)

Page 10: The Principle of Automatic Control 自动控制原理

10

The Principle of Automatic Control 2008

h(t)

t

时间 tr

上 升

峰值时间 tp

A

B

超调量 σ% = AB

100%

A typical unit-step response curve

调节时间 ts

误差带

稳态误差t

Page 11: The Principle of Automatic Control 自动控制原理

11

The Principle of Automatic Control 2008

动态性能指标• Rise time tr (上升时间) is defined as the time req

uired for the step response to rise from zero to its first steady-state value, or to rise from 10 to 90 percent of its steady-state value.

• Delay time td (延迟时间) is defined as the time required for the step response to reach 50 percent of its final value.

• Peak time tp (峰值时间) is defined as the time required for the step response to reach the first peak value.

• Setting time ts (调节时间) is defined as the time required for the step response to decrease and stay within a specified percentage (5% or 2%) of its final value.

Page 12: The Principle of Automatic Control 自动控制原理

12

The Principle of Automatic Control 2008

• Percent maximum overshoot σ% (超调量),响应的最大偏差与终值的查与终值之比的百分数

Note: The maximum overshoot is often used to measure the relative stability of a control system. A system with a large overshoot is usually undesirable.

%100)(

)()(%

h

hth p

实际应用中,常用的动态指标:上升时间,评价系统的响应速度( quickness)超调量,评价系统的阻尼程度 (smoothness)调节时间,同时反映响应速度和阻尼程度

Page 13: The Principle of Automatic Control 自动控制原理

13

The Principle of Automatic Control 2008

稳态性能指标

• Steady-state error (稳态误差) It is defined as the discrepancy between t

he output and the reference input when the steady state is reached.

• Detailed in 3-6

Page 14: The Principle of Automatic Control 自动控制原理

14

The Principle of Automatic Control 2008

Followed by…

3-2 Time response of a first-order system (一阶系统)

3-3 Transient response (瞬态响应) of a prototype 2nd-order system ( 2 阶原型系统)

Why we emphasize the 1-th order and 2-nd order systems?

• Higher-order systems can be considered to be sum of the responses of first- and second-order systems

Page 15: The Principle of Automatic Control 自动控制原理

15

The Principle of Automatic Control 2008

3-2 Time response of a first-order system (一阶系统)

Page 16: The Principle of Automatic Control 自动控制原理

16

The Principle of Automatic Control 2008

First-order system

0

0

( )( )

( )

bC sG s

R s s a

Time constant ( 时间常数)

Gain增益

1

K

Ts

Example

( )V s 1

sC

R

( )cV s

( )( )

( )cV s

G sV s

1

1sC

RsC

1

1RCs

Q : What is RC circuit used for?

Page 17: The Principle of Automatic Control 自动控制原理

17

The Principle of Automatic Control 2008

Time response( )

( )( )

C sG s

R s

1

1Ts

1. 零初始条件下系统的单位阶跃响应1

If ( ) 1( ) ( ) ,

1 1 1then ( ) ( ) ( )

1 1

r t t R ss

TC s G s R s

Ts s s Ts

/( ) 1 t Th t e Taking the inverse Laplace transform on both sides, it follows that

T时间常数

Page 18: The Principle of Automatic Control 自动控制原理

18

The Principle of Automatic Control 2008

0

0. 632

T t

C( t )

1

1/ T斜率 Tteth /1)( , ( ) 0.632

2 , (2 ) 0.865

3 , (3 ) 0.950

4 , (4 ) 0.982

t T h T

t T h T

t T h T

t T h T

/( ) 1 t Th t e

% 0 3 ( 5% ), 4 ( 2% )s st T t T 误差带 误差带

% ?

?

?

p

s

t

t

No overshoot, no oscillationpt 不存在

Page 19: The Principle of Automatic Control 自动控制原理

19

The Principle of Automatic Control 2008

一阶系统单位阶跃响应的特点:

2. 可用时间常数 T 度量系统输出量的数值

3. 响应曲线斜率初始值为 1/T ,可用来实验测定一阶系统参数 T

4. 时间参数反映了系统的惯性, T 越大系统响应越慢,惯性越大。

1. 响应曲线具有非振荡特征,故也称为非周期响应。

Page 20: The Principle of Automatic Control 自动控制原理

20

The Principle of Automatic Control 2008

Example 1. Consider a first-order system, (1) calculate the settling time ts when kt=0.1;(2) calculate the value of Kt that makes ts=0.

100/ s

kt

R( s) C( s)

Page 21: The Principle of Automatic Control 自动控制原理

21

The Principle of Automatic Control 2008

Closed-loop TF1/100 /

( )100 0.01

1 1

t

tt

ksG s

sk

s k

0.013 3 0.1,

0.3

st

t

t Tk

k

100/ s

kt

R( s) C( s)

100.1 , ( )

0.1 1tk G ss

当 时 0.1 .T 时间常数 秒

3 0.3st T 调节时间为 秒0.1st 要求 秒

Page 22: The Principle of Automatic Control 自动控制原理

22

The Principle of Automatic Control 2008

2

1( ) , ( )R t t R s

s

2. 零初始条件下系统的单位斜坡响应

/

/

Via inverse Laplace transform, unit ramp response is

( ) ( ) ( 0)

where is steady-state part being transient one

t Tt

t T

C t t T Te t

t T Te

, 。

2

2 2

1 1 1( ) ( ) ( )

1 1

T TC s G s R s

Ts s s s Ts

Page 23: The Principle of Automatic Control 自动控制原理

23

The Principle of Automatic Control 2008

The unit-ramp response is shown as follows :

Definition of steady-state error :When t tends to infinity, the difference between actual steady-state value and the expected one :

0 ( )

- ( - )sse h h

t t T

T

t

T

T

r(t)=t

c(t)

0

/( ) ( ) t TtC t t T Te

一阶系统单位斜坡达到稳态时具有和输入相同的斜率,只要在时间上滞后 T,这就存在着 ess=T 的稳态误差。

Page 24: The Principle of Automatic Control 自动控制原理

24

The Principle of Automatic Control 2008

课后复习• 自学中文教材 3 - 2 节

– 一阶系统单位脉冲响应– 一阶系统单位加速度响应

Page 25: The Principle of Automatic Control 自动控制原理

25

The Principle of Automatic Control 2008

3-3 Time response of a prototype

2-order system

Page 26: The Principle of Automatic Control 自动控制原理

26

The Principle of Automatic Control 2008

Second-order system

02

1 0

( )( )

( )

bC sG s

R s s a s a

Damping ratio阻尼比

Nature Frequency自然频率

2

2 22n

n ns

2 22 0n ns 特征多项式:2

1,2 1n ns 特征根:

二阶系统响应特性决定于根的位置,即阻尼比 ζ 和自然频率 ωn 这两个参数

Page 27: The Principle of Automatic Control 自动控制原理

27

The Principle of Automatic Control 2008 二阶系统单位阶跃响应

定性分析

2

2 2( )

2n

n n

G ss

ζ > 1

21,2 1n ns w jw

j

0

j

0

21,2 1n ns

0> ζ <1

ζ = 0

过阻尼

ζ =1 1,2 ns

j

0

临界阻尼欠阻尼

1,2 ns jw

j

0零阻尼

21,2 1n ns

Page 28: The Principle of Automatic Control 自动控制原理

28

The Principle of Automatic Control 2008

二阶系统的阻尼特性1. 当 0< ζ <1 时,一对负实部的共轭复根 , 系统的

单位阶跃响应具有衰减振荡特性,称为欠阻尼 (underdamped) 。

2.当 ζ =1 时,两个相等的负实根,称为临界阻尼 (critically damped)

3.当 ζ >1 时,特征方程具有两个不相等的负实根,称为过阻尼 (overdamped)

4.当 ζ =0 时,系统有一对共轭纯虚根,系统单位阶跃响应作等幅振荡,称为无阻尼 (undamped) 。

5.当 ζ < 0 时,此时系统特征方程具有一对正实部的共轭复根或正实根,系统的单位阶跃响应具有发散特性,称为 negatively damped).

Page 29: The Principle of Automatic Control 自动控制原理

29

The Principle of Automatic Control 2008

二阶系统的单位阶跃响应1. ζ >1 过阻尼系统

)1)(1(

1

)/1)(/1(

/1

2)(

)(

21

21

21

22

2

sTsT

TsTs

TT

wsws

w

sR

sC

nn

n

21

21

2

2

2

2

1

1

)1(1

);1(1

TTTT

w

wT

wT

n

nn

且设

Page 30: The Principle of Automatic Control 自动控制原理

30

The Principle of Automatic Control 2008

ssR

1)( 1 2

1 2

1/ 1( )

( 1/ )( 1/ )

T TC s

s T s T s

21 /

21

/

12

1

1/

1

1/

11)]([)( TtTt e

TTe

TTsCLth

h( t )

t0

1 起始速度小,然后上升速度逐渐加大,到达某一值后又减小,响应曲线不同于一阶系统。过阻尼二阶系统的动态性能指标主要是调节时间ts ,根据公式求 ts 的表达式很困难,一般用计算机计算出的曲线确定 ts 。

Page 31: The Principle of Automatic Control 自动控制原理

31

The Principle of Automatic Control 2008

2. ζ = 1 临界阻尼系统

nn

n

n

n

ssssssc

111)( 22

2

tn

etth n

)1(1)(

2( ) tn

n

dh tte

dt

0

响应曲线的变化率始终为正, t趋于无穷时,变化率趋于 0 。

临界阻尼系统单位阶跃响应是稳态值为 1 的无超调单调上升过程 (monoganly

Page 32: The Principle of Automatic Control 自动控制原理

32

The Principle of Automatic Control 2008

21 TT 1

175.4 Tts

21 4TT 1.25

13.3 Tts

1 24T T1.25

13st T

大阻尼二阶系统可近似等效为一阶系统,调节时间可用3T1 来估算。 WHY过(临界)阻尼二阶系统调节时间特性

Page 33: The Principle of Automatic Control 自动控制原理

33

The Principle of Automatic Control 2008

the eigenvalues of the closed-loop system are:

21,2 1n n ds w jw jw

3. 0< ζ <1 欠阻尼系统

2

damping constant

1

damping frequency

n

d n

w

w w

衰减系数

阻尼振荡频率

Im

Re

d

d

n

21arctan

Page 34: The Principle of Automatic Control 自动控制原理

34

The Principle of Automatic Control 2008

When r(t)=1(t), the system output is

2

Taking inverse Laplace transform, one gets

( ) 1 (cos sin )1

nw td dh t e w t w t

2

2

11 sin( )

1

w tn

d

ew t arctg

2

2 2

2 2 2 2

1( )

2

1

( ) ( )

n

n n

n n

n d n d

wC s

s w s w s

s w w

s s w w s w w

Page 35: The Principle of Automatic Control 自动控制原理

35

The Principle of Automatic Control 2008 h(t)

t0

1

2( ) 1 sin( )

1d

w tn

h t te

Steady value=1 , indicating that the steady state error=0 for unit step input signal

Transient part , damping sinusoidal term

Note: 欠阻尼二阶系统的单位阶跃响应曲线是按指数规律衰减到稳定值的,衰减速度取决于特征值实部 -ξwn 的大小,而衰减振荡的频率,取决于特征根虚部 wd 的大小。

Page 36: The Principle of Automatic Control 自动控制原理

36

The Principle of Automatic Control 2008

4. ζ =0 零阻尼系统

0( ) 1 sin( 90 ) 1 cos , ( 0)n nh t t t t

等幅振荡曲线,振荡频率为 wn

因此, wn 称为无阻尼振荡频率。

2( ) 1 sin( )

1

w tn

dh t te

21d n 21

arctg

Page 37: The Principle of Automatic Control 自动控制原理

37

The Principle of Automatic Control 2008

Effects of damping ratio

rise time rt

Overshoot pMThe oscillation is smaller

The speed of the response is slower

We are confronted with a necessary compromise between the speed of response and the allowable overshoot.

(for a given )n

Page 38: The Principle of Automatic Control 自动控制原理

38

The Principle of Automatic Control 2008

Followed by…Note: In control engineering, except those

systems that do not allow any oscillation,

usually a control system is desirable with

- moderate damping (allowing some

overshoot)

- quick response speed

- short settling timeTherefore, a second-order control system is usually designed as an underdamped system.

Page 39: The Principle of Automatic Control 自动控制原理

39

The Principle of Automatic Control 2008 underdamped system 欠阻尼二阶系

,

, n

阻尼自然振荡频率

2

2 2( )

2n

n n

G ss

21arctan

arccos

Im

Re

d

d

n

21,2 1n n ds w jw jw

2

,

1 ,

n

d n

w

w w

衰减系数

阻尼振荡频率

Page 40: The Principle of Automatic Control 自动控制原理

40

The Principle of Automatic Control 2008

Performance analysis

2( ) 1 sin( ) 1

1d

tn

y t te

,0<

2 0, ( 0, 1, 2, )

1

n rt

d r

et n n

2( ) 1, that is, 1 sin( ) 1

1

n rw t

r d r

ey t w t

1. Rise Time

2 sin( ) 0

1

n rt

d r

et

tr is the time needed for the response to reach the steady-state value for the first time, so n=1.

2

1r

d n

t

For a given wn , ζ ↓ , tr ↓ ; For a given ζ , wn↑ , tr ↓ .

Unit-step response:

Page 41: The Principle of Automatic Control 自动控制原理

41

The Principle of Automatic Control 2008

2 . Peak time

( )0

dy t

dt

22

2

2

( )( 1 ) sin

1

sin1

n

n

tnn d

tnd

dy te t

dt

e t

sin 0 ( 0, 1, 2, )d p d pt t n n

=0

tp is the time needed for the response to reach the maximum value for the first time, so n=1.

21d pp

d n

t t

For a given wn , ζ ↓ , tp ↓ ; For a given ζ , wn↑ , tp ↓ .

2( ) 1 sin( ) 1

1d

tn

y t te

,0<

Page 42: The Principle of Automatic Control 自动控制原理

42

The Principle of Automatic Control 2008

3. Overshoot

21p

d n

tw w

2( ) 1 sin( )

1

n pw t

p

ey t

( ) ( )% 100%

( )py t y

y

2

2

/

2

1

1

/

sin( ) sin 1

( ) 1

Suppose that ( ) 1

Th % 1us 00%

py t e

y

e

Im

Re

d

d

n

21arctan

Overshoot is a function of damping ratio ζ , independent of wn.

2( ) 1 sin( ) 1

1d

tn

y t te

,0<

Page 43: The Principle of Automatic Control 自动控制原理

43

The Principle of Automatic Control 2008

Damping ratioξ

ζ 增大, σ % 减小,通常为了获得良好的平稳性和快速性,阻尼比 ζ 取在 0.4-0.8 之间 , 相应的超调量 25%-2.5% 。

Page 44: The Principle of Automatic Control 自动控制原理

44

The Principle of Automatic Control 2008

根据定义:

不易求出 ts ,但可得出 wnts与 ζ 的关系曲线:

2

2sin( 1 ) 0.05 0.02

1

nw t

n s

ew t

4

调节时

2( ) 1 sin( ) 1

1d

w tn

h t w te

0<,

Page 45: The Principle of Automatic Control 自动控制原理

45

The Principle of Automatic Control 2008

Page 46: The Principle of Automatic Control 自动控制原理

46

The Principle of Automatic Control 2008

调节时间不连续的示意图

ζ 值的微小变化可引起调节时间 ts 显著的变化。

Page 47: The Principle of Automatic Control 自动控制原理

47

The Principle of Automatic Control 2008

当 ζ =0.68 ( 5% 误差带)或 ζ =0.76 ( 2%误差带),调节时间 ts 最短。所以通常的控制系统都设计成欠阻尼的。

曲线的不连续性,是由于 ζ 值的微小变化可引起调节时间显著变化而造成的。

Approximated calculation: 常用阻尼正弦振荡的包络线衰减到误差带之内所需时间来确定 ts 。

当 ζ <=0.8 时 , 常把 这一项

去掉。写成 即

21

sntwe

)1sin( 2 sntw

21 sntwe

Page 48: The Principle of Automatic Control 自动控制原理

48

The Principle of Automatic Control 2008

Note: 在设计系统时 , ζ 通常由要求的最大超调量决定 , 而调节时间则由无阻尼振荡频率 wn 来决定。

3 3 ( 5%)

4 4 ( 2%)

sn

sn

tw

tw

可近似表示为:

两边取对数 , 得:2

1 1ln( )

1s

n

tw

调节时间与闭环极点实部数值成反比,即极点离虚轴越远,调节时间越短。

Page 49: The Principle of Automatic Control 自动控制原理

49

The Principle of Automatic Control 2008

Oscillation times N Definition of N: 在调节时间内 , 响应曲线穿

越其稳态值次数的一半。

Td 为阻尼振荡的周期。

21

22,

nd

dd

s

wwT

T

tN

5

振荡次

)sin(1

1)(2

twe

thd

twn

Page 50: The Principle of Automatic Control 自动控制原理

50

The Principle of Automatic Control 2008

2

1r

d n

tw w

21p

d n

tw w

2/ 1 % 100%e

3( 5%)

4( 2%)

sn

sn

tw

tw

2

2 2,

1s

dd d n

tN T

T w w

标准二阶系统的性能指标

2

2 2( )

2n

n n

G ss

指标之间有矛盾,响应速度和阻尼程度不能同时达到最好结果

如何选取 ζ 和 wn 来满足系统设计要求:(1) 当 wn 一定,要减小 tr 和 tp ,必须减少 ζ 值,要减少 ts则应增大 ζ wn 值,而且 z 值有一定范围,不能过大。 (2) 增大 wn ,能使 tr , tp 和ts都减少。 (3) 最大超调量 σ% 只由 ζ决定 , ζ 越小, σ% 越大。所以,一般根据 σ% 的要求选择 ζ 值,在实际系统中, ζ 值一般在 0.5~ 0.8 之间 .

Page 51: The Principle of Automatic Control 自动控制原理

51

The Principle of Automatic Control 2008

二阶系统工程最佳参数在某些控制系统中常常采用所谓二阶工程最佳参数作为设计控制系统的依据。这种系统选择的参数使

707.02

1

系统单位阶跃响应暂态特性指标为:

最大超调量 上升时间

调节时间 ts ( 2% ) =8.43T (用近似公式求得为 8T ) ts ( 5% ) =4.14T (用近似公式求得为 6T )

%3.4%100%21

e

Ttn

r 7.41 2

Page 52: The Principle of Automatic Control 自动控制原理

52

The Principle of Automatic Control 2008

5

( 34.5)AK

s s -

R(s) C(s)

Example 1: Consider the following unit-feedback system

System input is the unit-step function, When the amplifier gains are KA=200, KA=1500, KA=13.5 respectively, can you calculate the time-domain specifications of the unit-step response ?

Investigate the effect of the amplifier gain KA on the system response

Page 53: The Principle of Automatic Control 自动控制原理

53

The Principle of Automatic Control 2008

Solution: The closed-loop transfer function is

2 1000, 2 34.5

34.531.6( / ), 0.545

2

n n

nn

w w

w rad sw

2

2

5( )( )

1 ( ) 34.5 5

1000200, ( )

34.5 1000

A

A

A

KG ss

G s s s K

K ss s

Page 54: The Principle of Automatic Control 自动控制原理

54

The Principle of Automatic Control 2008

According to the formula to calculate the performance indices, it follows that

2

2

/ 1

2

0.12(sec)1

30.174(sec)

% 100% 13%

10.72( )

2 2

p

n

sn

s ns

d

tw

tw

e

t wtN times

w

Page 55: The Principle of Automatic Control 自动控制原理

55

The Principle of Automatic Control 2008

由此可见 ,KA 越大 , ξ 越小 ,wn 越大 ,tp 越小 ,б% 越大 , 而调节时间 ts 无多大变化。

If 1500, then 86.2( / ); 0.2

0.037( ), 0.174( ), % 52.7%, 2.34( )A n

p s

K w rad s

t s t s N times

If 200, then 34.5( / ); 0.545

0.12( ), 0.174( ), % 13%, 0.72( )A n

p s

K w rad s

t s t s N times

KA=1500 。。。。

KA=13.5 。。。。

1.2),/(22.8,5.13When sradwK nA

过阻尼

Page 56: The Principle of Automatic Control 自动控制原理

56

The Principle of Automatic Control 2008

系统工作在过阻尼状态 , 峰值时间 , 超调量和振荡次数不存在 ,而调节时间可将二阶系统近似为大时间常数T的一阶系统来估计,即 :

调节时间比前两种 KA 大得多 ,虽然响应无超调 ,但过渡过程缓慢 , 曲线如下:

)1(1

)1(1

)(46.13

2

2

2

1

1

n

n

s

wT

wT

Tt 秒

1.2),/(22.8,5.13When sradwK nA

Page 57: The Principle of Automatic Control 自动控制原理

57

The Principle of Automatic Control 2008

0 2 4 6 8 10 12 14 16 18

0.2

0.4

0.6

0.81.0

1.21.4 )1500(2.0 AK

)200(545.0 AK

)5.13(1.2 AK

twn

)(tc

KA 增大, tp 减小, tr 减小,可以提高响应的快速性,但超调量也随之增加,仅靠调节放大器的增益,即比例调节,难以兼顾系统的快速性和平稳性,为了改善系统的动态性能,可采用比例-微分控制( PD) 或速度反馈控制,即对系统加入校正环节。

Page 58: The Principle of Automatic Control 自动控制原理

58

The Principle of Automatic Control 2008

课后自学• 中文教材 p91 ,过阻尼二阶系统动态分析• P93 ,二阶系统的单位斜坡响应

Page 59: The Principle of Automatic Control 自动控制原理

59

The Principle of Automatic Control 2008

Improve the 2nd system二阶系统性能的改善

Ref inputcontroller

Manipulated var.

plant

disturbanceOutputerror

sensor

5

( 34.5)AK

s s -

R(s) C(s)AK

5

( 34.5)s s

设计控制器,改变闭环系统的性能

回顾

Page 60: The Principle of Automatic Control 自动控制原理

60

The Principle of Automatic Control 2008

Example 2. 下图表示引入了一个比例微分控制的二阶系统 , 系统输出量同时受偏差信号

和偏差信号微分 的双重控制。试分析比例微分校正对系统性能的影响。

)(t)(t

)2(

2

n

n

wss

w

1

Tds

r ( t ) c( t )- +

)(t

)(t)(t

PI CONTROL

( )t 反映了误差变化的趋势,在出现位置误差之前,提前产生修正作用,从而改善系统性能。

Page 61: The Principle of Automatic Control 自动控制原理

61

The Principle of Automatic Control 2008

Open-loop transfer function :

( )G s

2 2

2 2 2 2 2

( 1) ( 1)

2 2n d n d

n n d n d n n

w T s w T s

s w s w T s w s w s w

Closed-loop transfer function :

ndd wT2

1

The equivalent damping ratio :

( 1),

2( 1)2

d n

n

k T s wk

ss

w

( ) ( )c pG s G s2 ( 1)

( 2 )n d

n

w T s

s s w

( )( )

1 ( )

G ss

G s

是二阶系统;不是二阶原型系统

开环增益

Page 62: The Principle of Automatic Control 自动控制原理

62

The Principle of Automatic Control 2008

比例微分控制使系统增加了一个闭环零点 s=-1/Td,前面给出的计算动态性能指标的公式不再适用。

由于稳态误差与开环增益成反比 , 因此适当选择开环增益和微分器的时间常数 Td, 即可减小稳态误差 , 又可获得良好的动态性能。

d 闭环系统阻尼比↑

系统动态过程的超调量

wn

不变 后面研究增加的零点对系统性能的影响

Page 63: The Principle of Automatic Control 自动控制原理

63

The Principle of Automatic Control 2008

Velocity feedback• Example 3.图 :

采用了速度反馈控制的二阶系统 ,试分析速度反馈校正对系统性能的影响。

R( s) c( s)

- -

kt s

)2(

2

n

n

wss

w

)(s

Why叫速度反馈?和 PD 控制区别?

Page 64: The Principle of Automatic Control 自动控制原理

64

The Principle of Automatic Control 2008

)2()2(

1

)2()(

2

2

2

2

tnn

n

n

tn

n

n

kwwss

w

wssskw

wssw

sG

• Open-loop transfer function

R( s) c( s)

- -

kt s

)2(

2

n

n

wss

w

)(s

Page 65: The Principle of Automatic Control 自动控制原理

65

The Principle of Automatic Control 2008

2( 1)2 n t n

ks

sw k w

不引入速度反馈开环增益

Note: k有所减小 , 增大了稳态误差 , 因此降低了系统的精度。

)2( nt

nwk

wk

2nw

k

2

2( )

( 2 )n

n n t

wG s

s s w w k

为系统的开环增益。

Page 66: The Principle of Automatic Control 自动控制原理

66

The Principle of Automatic Control 2008

Closed-loop transfer function:

Note: 显然 ,所以速度反馈同样可以增大系统的阻尼比 , 而不改变无阻尼振荡频率 wn, 因此 ,速度反馈可以改善系统的动态性能。

22

2

22

2

222

2

2)21

(2

2)(1

)()(

nnt

n

nntn

n

ntnn

n

wsws

w

wswkws

w

wskwsws

w

sG

sGs

t

ntt wk2

1Equivalent damping ratio :

Page 67: The Principle of Automatic Control 自动控制原理

67

The Principle of Automatic Control 2008

Summary for design: 在应用速度反馈校正时 , 应适当增大原系统的开环增益 , 以补偿速度反馈引起的开环增益减小 , 同时适当选择速度反馈系数 kt,使阻尼比 ξ t 增至适当数值 , 以减小系统的超调量 ,提高系统的响应速度 ,使系统满足各项性能指标的要求。

On-class exercise: 7-13 (English textbook pp.311)

Page 68: The Principle of Automatic Control 自动控制原理

68

The Principle of Automatic Control 2008

Followed by…

•前面研究了标准二阶系统的动态性能,对于有附加零点的系统,以及附加极点的系统(三阶及高阶系统),如何在此基础上研究系统的动态性能?

Page 69: The Principle of Automatic Control 自动控制原理

69

The Principle of Automatic Control 2008

3-4 Analysis of higher-order system

Page 70: The Principle of Automatic Control 自动控制原理

70

The Principle of Automatic Control 2008

3-order system2

02

0

( )( )( 2 )

n

n n

ss

s s s s

单位阶跃输入, ζ<1

*0 1 1

1( )

A B CC s

s s s s s s s

( )c t 1 0s tAe *( )c t

Page 71: The Principle of Automatic Control 自动控制原理

71

The Principle of Automatic Control 2008 实极点对系统响应的影响

j

0

结论 1 : 增加极点增加了阻尼

结论 2 :

增加的极点越靠近原点增加阻尼的作用越明显

超调↓上升时间,峰值时间,调节时间↑

结论 3 : 实极点比复极点更右时,明显过阻尼特性(趋向于 1 阶响应)

Page 72: The Principle of Automatic Control 自动控制原理

72

The Principle of Automatic Control 2008

Higher-order systems

高阶系统的闭环传递函数可表示为如下普通形式

将分子和分母分解成因式,上式可写成

10 1 1

10 1 1

( )( )

( )

m mm m

n nn n

b s b s b s bC ss

R s a s a s a s a

1 2

1 2

( )( ) ( )( )

( )( ) ( )m

n

K s z s z s zs

s p s p s p

线性定常系统,零极点只可能是实数或共轭复数

Page 73: The Principle of Automatic Control 自动控制原理

73

The Principle of Automatic Control 2008

1

2 2

1 1

( )1

( )( ) ( 2 ))

m

ii

q r

j k k kj k

K s zC s

ss p s s

02 2

1 1

( )2

q rj k k

j kj k k k

AA B s CC s

s s p s s

20

1 1

2

21

( ) cos( 1 )

sin( 1 )1

j k k

k k

q rp t t

j k k kj k

rtk k k k

k kk k k

c t A A e B e t

C Be t

Page 74: The Principle of Automatic Control 自动控制原理

74

The Principle of Automatic Control 2008

20

1 1

2

21

( ) cos( 1 )

sin( 1 )1

j k k

k k

q rp t t

j k k kj k

rtk k k k

k kk k k

c t A A e B e t

C Be t

一阶响应函数Exponential curves

二阶响应函数Damped sinusoida

l curves+

如果所有闭环极点都在左半 s平面,随着 t增大,一阶指数及二阶衰减正弦分量趋于 0 ,高阶系统是稳定的,稳态量为A0

极点负实部越大(离虚轴远),响应分量衰减越快,对系统性能影响越小,反之越靠近虚轴,响应分量衰减越慢,对系统响应形式影响越大。

Page 75: The Principle of Automatic Control 自动控制原理

75

The Principle of Automatic Control 2008

Effects of Adding Poles and Zeros to Transfer Functions

Page 76: The Principle of Automatic Control 自动控制原理

76

The Principle of Automatic Control 2008

( )R s ( )Y s2

( 2 )n

ns s

1

1 pT s

( )R s ( )Y s2

( 2 )n

ns s

Closed-loop TF :

2

2 2

( ) ( )( )

( ) 1 ( ) 2n

n n

Y s G ss

R s G s s s

1 ps T

2

( )( 2 (1 ))

n

n p

G ss T ss

Open-loop TF :

Open-loop TF :2

( )( 2 )

n

n

G ss s

Closed-loop TF :2

3 2 2( )

(1 2 ) 2n

p n p n n

sT s T s s

Effects of Adding Poles

Page 77: The Principle of Automatic Control 自动控制原理

77

The Principle of Automatic Control 2008 1 ps T

0, 1, 2, 5pT : , 1, 0.5, 0.2poles s

j

01

0.5

0.2

-- Increasing the maximum overshoot of the closed-loop system;

How does it affect closed-loop system step-response performance?

-- Increasing the rise time of the closed-loop system.

Page 78: The Principle of Automatic Control 自动控制原理

78

The Principle of Automatic Control 2008

( )R s ( )Y s2

( 2 )n

ns s

Closed-loop TF :

Open-loop TF :

2

2 2

( ) ( )( )

( ) 1 ( ) 2n

n n

Y s G ss

R s G s s s

Closed-loop TF:

2

( )( 2 )

n

n

G ss s

1 ps T

Effects of Adding Poles

2

2 2( )

( (12 )) p

n

n n T ss

s s

2

3 2 2 2(1 2 ) (2 )n

p n p n n p nT s T s T s

Page 79: The Principle of Automatic Control 自动控制原理

79

The Principle of Automatic Control 2008 1 ps T

0, 0.5, 1, 2, 5pT

, 2, 1, 0.5, 0.2s

j

01

0.5

0.22

-- the maximum overshoot of the closed-loop system decreases;-- the rise time of the closed-loop system increases.

As the pole at is moved toward the origin in the s-plane1 ps T

opposite effects to that of adding a pole to

open-loop TF

1. Adding a pole to the closed-loop system has the effect as increasing the damping ratio;

2. An originally underdamped system can be made into overdamped by adding a closed-loop pole.

Page 80: The Principle of Automatic Control 自动控制原理

80

The Principle of Automatic Control 2008

( )R s ( )Y s2

( 2 )n

ns s

Closed-loop TF :

Open-loop TF :

2

2 2

( ) ( )( )

( ) 1 ( ) 2n

n n

Y s G ss

R s G s s s

1 zs T

Closed-loop TF:

2

( )( 2 )

n

n

G ss s

2

2 2

(1( )(

2

))

( )n

n n

zTY ss

R

s

s s s

Effects of Adding Zeros

2 2

2 2 2 22 2n n

n n n n

z

s s s

s

s

T

Page 81: The Principle of Automatic Control 自动控制原理

81

The Principle of Automatic Control 2008 1 zs T

2

2 2

(1( )(

2

))

( )n

n n

zTY ss

R

s

s s s

2

2 2

2

2 2 22n

n

n z

n nn

s

s ss s

T

For a unit-step input ,

1( )t 1( )R s

s

The step response of the closed-loop system22

2 2 22

1 1( ) ( ) ( )

2 2n

n n

nz

n n

Y s s R ssss s

Tss

s

1 11

( )( ) [ ( )] ( ) z

dy ty t L Y s y t T

dt

Closed-loop TF:

Page 82: The Principle of Automatic Control 自动控制原理

82

The Principle of Automatic Control 2008 1 zs T

0, 1, 3, 6, 10ZT

1 1 1, 1, , ,

3 6 10s

j

01 1

3

1

6

1

10

-- the maximum overshoot of the closed-loop system increases;

-- the rise time of the closed-loop system decreases.

As the zero at is moved toward the origin in the s-plane1 zs T

a) Its effects on an underdamped ( )system 0 1

The additional zero has the effect as reducing the damping ratio

Page 83: The Principle of Automatic Control 自动控制原理

83

The Principle of Automatic Control 2008 1 zs T

0, 1, 3, 6, 10ZT

1 1 1, 1, , ,

3 6 10s

j

01 1

3

1

6

1

10

b) Its effects on an overdamped ( )system 1

Adding a zero to an overdamped system can change it into an underdamped system by putting the zero at a proper position.

Page 84: The Principle of Automatic Control 自动控制原理

84

The Principle of Automatic Control 2008

( )R s ( )Y s2

( 2 )n

ns s

1 zT s

( )R s ( )Y s2

( 2 )n

ns s

Closed-loop TF :

Open-loop TF :

2

2 2

( ) ( )( )

( ) 1 ( ) 2n

n n

Y s G ss

R s G s s s

1 zs T

2

( )( 2 )

(1 )z n

n

T sG s

s s

Open-loop TF :

Closed-loop TF :

2

( )( 2 )

n

n

G ss s

2

2

2 2

1( )( )

(2 )z

z n

n

n n

ss

T s

T s

Effects of Adding Zeros

The additional zero changes

both numerator and

denominator.

Page 85: The Principle of Automatic Control 自动控制原理

85

The Principle of Automatic Control 2008 1 zs T

Closed-loop TF :2

2 2 2

(1 )( )

(2 )n z

n z n n

T ss

s T s

The additional zero changes

both numerator and

denominator.2

2 2

(1 )

2n z

n nz

T s

s s

2z n

z

T

The equivalent damping ratio:

zT z overshoot %

An additional zero will increase overshoot

1 zs T

zT overshoot %

1

overdampe

when , the closed-loop system

becomes no overshoot

no matter how large i ) s

(dz

zT

Page 86: The Principle of Automatic Control 自动控制原理

86

The Principle of Automatic Control 2008

Dominant Poles of Transfer Function

Dominant poles: those poses that have a dominant effect on the transient response.By identifying dominant poles, high-order systems can be approximated by lower ones as the transient response is concerned.

close to the imaginary axis

Position of Poles in the left-half s-plane Their effects on transient response

decaying relatively slowlyfar away from the imaginary

axis decaying fast

31 21 2 3( ) p tp t p ty t c e c e c e 1 2 3poles: - , - , -p p p

1 2 3If ,p p p 11 decays fastest,p tc e 3

3 decays slowest.p tc e

1 2 3

1( )

( )( )( )Y s

s p s p s p

e.g.31 2

1 3 3

cc c

s p s p s p

Page 87: The Principle of Automatic Control 自动控制原理

87

The Principle of Automatic Control 2008

87

115

If the ratio of real parts exceed 5 and no zeros nearby, the closed-loop poles nearest the imaginary-axis will dominate in the transient response behavior.

The dominant poles can be a real pole, but a pair of complex conjugate poles are more preferable in control engineering ( why? ) .

In order to apply second-order system in approximating the dynamic performance of higher-order system

Dominant Poles of Transfer Function

Page 88: The Principle of Automatic Control 自动控制原理

88

The Principle of Automatic Control 2008

3-5 Stability of linear systems (in complex plane)

Page 89: The Principle of Automatic Control 自动控制原理

89

The Principle of Automatic Control 2008

An example

a,b,c点,哪些是稳定 (stable) 的,哪些是不稳定 (unstable) 的?

a

b

c

b

Page 90: The Principle of Automatic Control 自动控制原理

90

The Principle of Automatic Control 2008

若系统在初始偏差作用下 ,其过渡过程随时间的推移 ,逐渐衰减并趋于零 , 具有恢复平衡状态的性能 ,则称该系统为 Asymptotic stable(渐进稳定),反之为 unstable 。

Note 1: For linear systems, stability means asymptotic stability, but it is not true for nonlinear systems.

Note 2: Stability of linear systems depends only on internal structure of the system, independent of external action and initial condition.

Definition

Page 91: The Principle of Automatic Control 自动控制原理

91

The Principle of Automatic Control 2008

)()(

)()( isoutput system The

)(

)(

)(

)(

)(

)()(

11

10

11

10

11

10

11

10

sRsD

sMsC

asasasasD

bsbsbsbsM

sD

sM

asasasa

bsbsbsb

sR

sCs

nnnn

mmmm

nnnn

mmmm

Sufficient and necessary condition (充分必要条件) for stability

Page 92: The Principle of Automatic Control 自动控制原理

92

The Principle of Automatic Control 2008

The initial condition of the system is zero. When system input is unit impulse functionδ(t), the system output is k(t).

If , then the system is stable 。0)( lim

tkt

n

i i

i

ss

C

sD

sMsC

1)(

)()(

时域定义:

n

i

tsi

ieCtctk1

)()(

0)( lim

tkt

decay with timeis te

All poles should locate in the left side of s-plane

Page 93: The Principle of Automatic Control 自动控制原理

93

Review of Chapter3

• 3-5 Stability of linear control systems

– Sufficient and necessary condition (充分必要条件) for stability

All poles should locate in the left side of s-plane

Page 94: The Principle of Automatic Control 自动控制原理

94

The Principle of Automatic Control 2008

A system is stable if and only if all roots

of the system characteristic equation

have negative real parts, or equivalently,

all poles of closed-loop transfer functions

must locate in the left half of s-plane.

Note: The above criterion holds when the characteristic equation has multiple-order roots 。

Stability criterion in complex plane

Page 95: The Principle of Automatic Control 自动控制原理

95

The Principle of Automatic Control 2008

Routh-Hurwitz stability criterion (劳思 -赫尔维茨稳定判据)

• The criterion tests whether any of the roots of the characteristic equation lie in the right half of the s-plane, without actually solving for the roots.

• Information about absolute stability can be obtained directly from the coefficients of the characteristic equation

All poles in

left s

plane

No poles in

right s plane=

Page 96: The Principle of Automatic Control 自动控制原理

96

The Principle of Automatic Control 2008

1.The necessary condition: Consider that the characteristic equation of a LTI system is of the form

The necessary (but not sufficient) conditions for stability of the system are described as follows:

(1) All the coefficients of the equation have the same sign. (a0>0, 各项系数为正)

(2) None of the coefficients vanishes.

10 1 1 0( ) ... 0, 0n n

n nD s a s a s a s a a

Procedures in Routh’s stability criterion

Page 97: The Principle of Automatic Control 自动控制原理

97

The Principle of Automatic Control 2008

Note 1: The above conditions are based on the laws of algebra.

Note 2: These conditions are not sufficient. (不是充分条件)

1 2 3( )( )( ) 0s s s s s s

2 3 1 2 3( )( ) ( )( ) 0s s s s s s s s s s 2 2

2 3 2 3 1 2 3 2 3( ( ) ) ( ( ) ) 0s s s s s s s s s s s s s s 3 2

1 2 3 1 2 1 3 2 3 1 2 3( ) ( ) 0s s s s s s s s s s s s s s s

3 20 1 2 3 0a s a s a s a

31

10i

i

as

a

3

2

, 10i j

i ji j

as s

a

3

3

, , 10i j k

i j ki j k

as s s

a

如果都是负实根,特征多项式所有系数均为正,否则必有正根或虚根

Page 98: The Principle of Automatic Control 自动控制原理

98

The Principle of Automatic Control 2008

2 Routh’s Tabulation (劳斯表) or Routh’s Array(劳斯矩阵)

若第 1步特征多项式系数不满足必要条件(所有系数不为正),系统一定不稳定,否则需要构造劳斯表进一步判断系统的稳定性

31

20

11

1aa

aa

ab

51

40

12

1aa

aa

ab

71

60

13

1aa

aa

ab

21

31

11

1bb

aa

bc

31

51

12

1bb

aa

bc

41

71

13

1bb

aa

bc

10

11

212

43213

43212

75311

6420

gs

fs

ees

ccccs

bbbbs

aaaas

aaaas

n

n

n

n

Page 99: The Principle of Automatic Control 自动控制原理

99

The Principle of Automatic Control 2008

• Routh criterion: • The system is stable if and only if all the

elements of the first column of the Routh’s tabulation are of the same sign.(表中第一列系数均为正数)

• The number of changes of signs in the elements of the first column equals the number of roots with positive real parts or in the right half of the s-plane. (符号改变次数等于右半平面根的个数)

Page 100: The Principle of Automatic Control 自动控制原理

100

The Principle of Automatic Control 2008

4

3

2

1

0

1 3 5

2 4

1 5

6 0

5

s

s

s

s

s

Example4 3 22 3 4 5 0s s s s

则系统不稳定 ,且有两个正实部根 , 即有 2 个根在 S的右半平面。

Page 101: The Principle of Automatic Control 自动控制原理

101

The Principle of Automatic Control 2008

一次方程 :

a1,a0 同号则系统稳定。

二次方程 :

a1,a2,a0 同号则系统稳定。

三次方程 :

0 1 0a s a

20 1 2 0a s a s a

3 20 1 2 3 0a s a s a s a

30 2

21 3

1 0 3 1 2

1

03

s a a

s a a

a a a as

a

s a

a0,a1,a2,a3均大于0,且 a1a2>a3a0, 则系统稳定。

Page 102: The Principle of Automatic Control 自动控制原理

102

The Principle of Automatic Control 2008

3. Special cases when Routh’s Tabulation terminates prematurely

Case 1: 劳斯表中某一行的第一个元素为 0, 其它各元素不全为 0

用任意小的正数 ε 代替某一行第一个为 0 的元素。然后继续劳斯表计算并判断。

16

04812

16)(0

123

1641

0161243

0

1

2

3

4

234

s

s

s

s

s

ssss

当 ε很小时 ,

则系统不稳定,并有两个正实部根。

048

124812

Page 103: The Principle of Automatic Control 自动控制原理

103

The Principle of Automatic Control 2008

Case 2:劳斯表中第k行元素全为 0, 这说明系统的特征根或存在两个符号相异 ,绝对值相同的实根 , 或存在一对共轭纯虚根 , 或存在实部符号相异 , 虚部数值相同的共轭复根,或上述类型的根兼而有之。

此时系统必然是不稳定的。在这种情况下 , 可作如下处理。

(1). 用 k-1行(上一行)元素构成辅助方程 (2).将辅助方程为 s 求导 , 其系数作为全零行的元素 ,继续完成劳斯表。

(3). 可通过解辅助方程获取不稳定根的信息

Page 104: The Principle of Automatic Control 自动控制原理

104

The Principle of Automatic Control 2008

Example : The characteristic equation of a system is : 0124933 2345 sssss

0

1

2

3

4

5

000

1293

431

s

s

s

s

s

s

01812 is derivative Its

012933

24

ss

ss

5

4

3

2

1

0

1 3 4

3 9 12

12 18 0

s

s

s

s

s

s

12

050

122/9

01812

1293

431

0

1

2

3

4

5

s

s

s

s

s

s

The sign in the first column changes once, so the system is unstable.

Page 105: The Principle of Automatic Control 自动控制原理

105

The Principle of Automatic Control 2008

辅助方程得:

解得符号相异,绝对值相同的两个实根

和一对纯虚根

其中有一个正实根导致系统不稳定。

0)4)(1(43 2224 ssss原式=

12,1 s

3,4 2s j

4 23 9 12 0s s

Page 106: The Principle of Automatic Control 自动控制原理

106

The Principle of Automatic Control 2008

On-class exercisesDetermine the stability of the

following systems:

0322 )2(

010532 )1(

234

234

ssss

ssss

Page 107: The Principle of Automatic Control 自动控制原理

107

The Principle of Automatic Control 2008

调节器 被控对象

检测元件

R( s) E( s)

-

N( s)扰动C( s)被调量

k

2

(1).劳斯表不但可判断系统的稳定性 ,而且能判断特征根的位置分布情况。

(2). 对于带调节器的系统,可应用劳斯判据选择使系统稳定的调节器参数或取值区间。

Example 1:

)2)(1(

10

sss

Application and extension of Routh Tabulation

K

2

Page 108: The Principle of Automatic Control 自动控制原理

108

The Principle of Automatic Control 2008

10( )( ) ( 1)( 2)

( )20( ) 1 ( ) 1

( 1)( 2)

forward

loop

kG sC s s s s

skR s G s

s s s

调节器 被控对象

检测元件

R( s) E( s)

-

N( s)扰动C( s)被调量

k

2

)2)(1(

10

sssK

2

1 ( ) 0

201 0

( 1)( 2)

G s

k

s s s

l oop特征方程:

Page 109: The Principle of Automatic Control 自动控制原理

109

The Principle of Automatic Control 2008

Characteristic equation

Necessary condition:

Sufficient condition: 1 2 3 0 6 20

0.3

a a a a k

k

201 0

( 1)( 2)

k

s s s

20 0 0k k

3 23 2 20 0s s s k

0 0.3 , the system is stablek

Page 110: The Principle of Automatic Control 自动控制原理

110

The Principle of Automatic Control 2008

Example 2:已知系统的特征为 :

试判断使系统稳定的 k 值范围 ,如果要求特征值均位于 s=-1垂线之左。问 k 值应如何调整 ?

0325.0025.0 23 ksss

Page 111: The Principle of Automatic Control 自动控制原理

111

The Principle of Automatic Control 2008

解 : 特征方程化为 : 列劳斯表 :

0404013 23 ksss

3

2

1

0

1 40

13 40

13 40 40

1340

s

s k

ks

s k

0325.0025.0 23 ksss

13

0

k

k

Page 112: The Principle of Automatic Control 自动控制原理

112

The Principle of Automatic Control 2008

使系统稳定的 k值范围是

若要求全部特征根在 s=-1 之左,怎么办?虚轴向左平移一个单位,令 s=s1-1代入原特征方程 ,

得 :

整理得 :

130 k

040)1(40)1(13)1( 12

13

1 ksss

0)2840(1710 121

31 ksss

0404013 23 ksss

Page 113: The Principle of Automatic Control 自动控制原理

113

The Principle of Automatic Control 2008

列劳斯表 :

第一列元素均大于 0, 则得 :

284010

)2840(170284010

171

01

11

21

31

ks

ks

ks

s

95.47.0 k

0)2840(1710 121

31 ksss

Page 114: The Principle of Automatic Control 自动控制原理

114

The Principle of Automatic Control 2008

3-6 Steady-state error

Page 115: The Principle of Automatic Control 自动控制原理

115

The Principle of Automatic Control 2008

Review

• Three basic requirements of control system:– Stability

– Quickness

– Accuracy

Locations of closed-loop poles

Time responses of 1,2 and higher order systems

1-order system, T 2-order system, ζωn

Higher order, dominant poles

steady-state error

Page 116: The Principle of Automatic Control 自动控制原理

116

The Principle of Automatic Control 2008

In a design problem, one of the objectives is to keep the actual output tracking the reference input accurately, that is, keep the steady-state error to a minimum, or below a certain tolerable value, and at the same time the transient response must satisfy a certain set of specifications.

Page 117: The Principle of Automatic Control 自动控制原理

117

The Principle of Automatic Control 2008

Error and steady-state error in control systems

)()()( tctcte r

Ref input

r(t)controller

Manipulated var.

plant

disturbanceOutput

c(t)

Error

sensor

Error:

Steady-state error:

)]()([lim)(lim tctctee rtt

ss

理想输出 cr

(t)

Page 118: The Principle of Automatic Control 自动控制原理

118

The Principle of Automatic Control 2008

(1) According to Final-value Theorem

Note: When input signals are

we can use Final-value Theorem to calculate steady-state error, not for sinusoid input signal (refer to final-value theorem).

0lim [ ( ) ( )]rs

s C s C s

2

2

1,),(1),( tttt

How to calculate steady-state error

0lim ( ) lim ( )t s

f t s F s

lim ( )sst

e e t

0

lim ( )s

s E s

Page 119: The Principle of Automatic Control 自动控制原理

119

The Principle of Automatic Control 2008

(2) According to the definitiona.求误差响应传递函数

b. 误差响应的象函数

c. 误差响应的原函数

d.求极值 即为稳态误差。

)(

)(

sR

sE

lim ( )ss te e t

)()(

)()( sR

sR

sEsE

)(te

)(lim tet

R(s)G(s)

C(s)E(s)

H(s)

Page 120: The Principle of Automatic Control 自动控制原理

120

The Principle of Automatic Control 2008

If there exist both control input signal and disturbance signal, then steady-state error can be calculated as follows: :

R(s)

N(s)

E(s)+

+

)(ser

)(sen

利用叠加原理:

)]()()()([lim)(lim00

sNssRssssEe enerss

ss

R(s)Gc(s)

C(s)E(s)

H(s)

N(s)

Gp(s)-

如何求误差输入传函和误差扰动传函?

Page 121: The Principle of Automatic Control 自动控制原理

121

The Principle of Automatic Control 2008

N(s)R(s)

Gc(s)C(s)E(s)

H(s)

Gp(s)-

系统在输入信号和扰动信号作用下的实际输出为 :

1 2

( ) ( ) ( ) ( ) ( )

( ) ( )cr cnC s s R s s N s

C s C s

系统对输入信号R(s) 的理想输出 Cr(s):

( ) 0 ( ) ( ) ( )

( ) ( )

( )

r

r

E s R s H s C s

R sC s

H s

系统对扰动信号 N(s) 的理想输出为 0 ,即不希望系统受到扰动的影响

( ) ( )( )

1 ( ) ( ) ( )c p

crc p

G s G ss

G s G s H s

( )( )

1 ( ) ( ) ( )p

cnc p

G ss

G s G s H s

Page 122: The Principle of Automatic Control 自动控制原理

122

The Principle of Automatic Control 2008

1 1( )

1 ( ) ( ) ( ) ( )erc p

sG s G s H s H s

在输入信号和扰动信号作用下的误差 E(s) 为 :

1 1

( ) ( )1( ) ( ) ( ) [ ] ( )

( ) 1 ( ) ( ) ( )

( ) ( )

c pr

c p

er

G s G sE s C s C s R s

H s G s G s H s

s R s

2 2

( )( ) 0 ( ) ( )

1 ( ) ( ) ( )

( ) ( )

p

c p

en

G sE s C s N s

G s G s H s

s N s

( )( )

1 ( ) ( ) ( )p

enc p

G ss

G s G s H s

1 2( ) ( ) ( ) ( ) ( ) ( ) ( )er enE s E s E s s R s s N s

Page 123: The Principle of Automatic Control 自动控制原理

123

The Principle of Automatic Control 2008

Example :已知系统的结构图如下 ,试求系统在输入信号 r(t)=t 和扰动信号 n(t)=-1(t) 同时作用下系统的稳态误差 ess

2

n(t)=-1(t)r(t)=t E(s)

-C(s)

12.0

1

s )1(

2

ss

Page 124: The Principle of Automatic Control 自动控制原理

124

The Principle of Automatic Control 2008

2

n(t)=-1(t)r(t)=t E(s)

-C(s)

12.0

1

s )1(

2

ss

2

1( ) ( )r t t R s

s

1( ) 1( ) ( )n t t N s

s

1( )

0.2 1cG ss

2( )

( 1)pG ss s

( ) 2H s

Page 125: The Principle of Automatic Control 自动控制原理

125

The Principle of Automatic Control 2008

At first, we check stability of the system 。 ( 稳定性只和系统内部结构有关,和外部输入及扰动无关)The characteristic equation is:

1 21 ( ) ( ) ( ) 1 2 0

0.2 1 ( 1)c pG s G s H ss s s

1( )

0.2 1cG ss

2( )

( 1)pG ss s

( ) 2H s

1 2 3 01.2 0.8 . ., i e a a a a 由劳斯判据

系统稳定

3 20.2 1.2 4 0s s s

特征方程

Page 126: The Principle of Automatic Control 自动控制原理

126

The Principle of Automatic Control 2008

2

( 1)1 2

1 20.2 1 ( 1)

s s

s s s

2

1( ) ( )r t t R s

s

1( ) 1( ) ( )n t t N s

s

1( )

0.2 1cG ss

2( )

( 1)pG ss s

( ) 2H s

1 1( )

1 ( ) ( ) ( ) ( )erc p

sG s G s H s H s

( )( )

1 ( ) ( ) ( )p

enc p

G ss

G s G s H s

( ) ( ) ( ) ( ) ( )er enE s s R s s N s 1 1

1 2 21 20.2 1 ( 1)s s s

0 lim ( )ss

se sE s

5

8

Page 127: The Principle of Automatic Control 自动控制原理

127

The Principle of Automatic Control 2008

Under only input signal, the system error is:

If a system is unity feedback ( 单位反馈H(s)=1)

1 1( ) ( )

1 ( ) ( ) ( ) ( )c p

E s R sG s G s H s H s

1( ) ( ) ( ) ( )

1 ( ) ( )c p

E s R s C s R sG s G s

Steady-state error of system with input signal

N(s)R(s)

Gc(s)C(s)E(s)

H(s)

Gp(s)-

Ess 和系统内部结构以及外部输入密切相关

开环传递函数

Page 128: The Principle of Automatic Control 自动控制原理

128

The Principle of Automatic Control 2008

ν为开环系统在 s平面原点的极点重数

Whenν=0,1,2, the system is called type 0, type 1, type 2; k is called open-loop gain.

2 21 2 1 2

0 02 21 2 2 2

( 1) ( 2 1)( ) , 0 , ( ) 1

( 1) ( 2 1)

s s sG s s G s

T s T s T s

当 时

0 0

0

1lim ( ) lim ( )

1 ( )ss

s se sE s s R s

kG s

s

0( ) ( ) ( ) ( )c p v

kG s G s G s G s

s

系统类型

为了进一步分析误差和系统结构的关系

0

1( ) ( )

1 ( )E s R s

kG s

s

Page 129: The Principle of Automatic Control 自动控制原理

129

The Principle of Automatic Control 2008

0 0

00

1lim ( ) lim ( )

1 ( )ss

s se sE s s R s

kG s

s

ν=0, type 0 system

阶跃输入, R(s)=1/s1

1ssek

斜坡输入, R(s)=1/s2sse

加速度输入, R(s)=1/s3sse

有稳态误差

不稳定

不稳定

Page 130: The Principle of Automatic Control 自动控制原理

130

The Principle of Automatic Control 2008

0 0

01

1lim ( ) lim ( )

1 ( )ss s s

e sE s s R sk

G ss

ν=1, type 1 system

阶跃输入, R(s)=1/s1

01sse

斜坡输入, R(s)=1/s21

ssek

加速度输入, R(s)=1/s3sse

无稳态误差

有稳态误差

不稳定

Type-1 system can track step signal accurately.

Page 131: The Principle of Automatic Control 自动控制原理

131

The Principle of Automatic Control 2008

0 0

02

1lim ( ) lim ( )

1 ( )ss

s se sE s s R s

kG s

s

ν=2, type 2 system

阶跃输入, R(s)=1/s1

01sse

斜坡输入, R(s)=1/s21

00sse

加速度输入, R(s)=1/s31

ssek

无稳态误差

有稳态误差

无稳态误差

Type-2 system can track step and ramp signals accurately.

Page 132: The Principle of Automatic Control 自动控制原理

132

The Principle of Automatic Control 2008

Define kp, kv, ka as steady-state error constants:Under step-function input - position error

constant ( 位置误差系数 )

Under ramp-function input - Velocity error

constant( 速度误差系数 )

Under parabolic-function input - Accerlation error

constant( 加速度误差系数 )

s

ksG

s

kk

ssp

00

0lim)(lim

10010

lim)(lim s

ksG

s

kk

ss

20020

lim)(lim s

ksG

s

kk

ssa

0 0

00

1lim ( ) lim ( )

1 ( )ss s s

e sE s s R sk

G ss

Page 133: The Principle of Automatic Control 自动控制原理

133

The Principle of Automatic Control 2008

Type ofSystem

Errorconstants

Steady-state error

j

0

)(1)( 0 tRtr tVtr 0)( 2)( 20tAtr

k

R

10

k

V0

k

A0

0

00

0

00k

k

k

pk vk ak

sse

Summary of steady-state error and error constants for unit-feedback systems (H(s)=1)

Page 134: The Principle of Automatic Control 自动控制原理

134

The Principle of Automatic Control 2008

应用误差常数 (kp,kv,ka) 计算稳态误差的几点说明

• 1.闭环系统一定是稳定的,否则没有意义。• 2. 上表中总结的结果仅针对单位反馈系统

有效。• 3. kp,kv,ka 等误差常数是分别针对阶跃、斜

坡、抛物线形式的输入定义的,当输入是这三种信号组合时,可应用叠加原理计算 ess

Page 135: The Principle of Automatic Control 自动控制原理

135

The Principle of Automatic Control 2008

• 4. 以开环系统在 s平面原点上极点重数定义系统类别的优点:– 可以根据输入信号形式,迅速判断系统是否存

在稳态误差以及稳态误差的大小• 5. 影响稳态误差的主要因素:

• 系统结构(系统类别,开环增益等)• 输入信号

Note 1: 提高系统的型别 , 增大系统的开环增益 ,都会提高系统的精度 , 但这样又会降低稳定性 ,必须综合考虑。

Page 136: The Principle of Automatic Control 自动控制原理

136

The Principle of Automatic Control 2008 Example 1:

已知单位反馈系统开环传递函数为 G(s) ,输入为 r(t) ,试求稳态误差 ess 。

s2(0.1s+1)8(0.5s+1)

G3(s)=

s(s+4)(s2+2s+2)7(s+3)

G2(s)=

(0.1s+1)(0.5s+1)10

G1(s)=r1(t)=1(t)

r2(t)=t

r3(t)=t2

解:

0型

Ⅰ型

Ⅱ型

k=10

k=21/8

k=8

ess=1/11

ess= 8/21

ess=1/8

×

×

系统 2 不稳定,系统 3 的 A=2 ,

∴ ess→∞

∴ ess=1/4

Page 137: The Principle of Automatic Control 自动控制原理

137

The Principle of Automatic Control 2008

On-class Exercise: Consider the unity feedback system with the following transfer functions.

Problem: Calculate error constants and steady-state

errors for the three basic types of inputs.

)5.0)(5.1(

)15.3()( )(

sss

sKsGa

)12()( )(

2

ss

KsGb

)5)(12(

)1(5)( )(

2

sss

ssGc

Page 138: The Principle of Automatic Control 自动控制原理

138

The Principle of Automatic Control 2008

Example: 某控制系统的结构图为

试分别求出 H(s)=1 和 H(s)=0.5 时系统的稳态误差。

12

102 ss

)(sH

)(sC

-

)(15)( ttr

可直接利用开环系统型别计算 ess

利用定义计算 ess

Page 139: The Principle of Automatic Control 自动控制原理

139

The Principle of Automatic Control 2008

解 :, 系统的开环传递函数为

当H(s)=1 时

当H(s)=0.5 时 ,

11

5

101

5

10

k

Ress

0lim ( ) ( )ss ers

e s s R s

12

102 ss

)(sH

)(sC

-

)(15)( ttr

?, ?k 2

10( )

2 1G s

s s

10, 0k

0

2

1 1 5 5lim

10 0.5 31 0.52 1

ss

ss s

Page 140: The Principle of Automatic Control 自动控制原理

140

The Principle of Automatic Control 2008

思考:(1) 若上例在 H(s)=1 时 , 系统的允许误差为 0.2,问开环增益 k应等于多少 ?

(2)当 时 , 上例的稳态误差又是多少 ?

0 0 5(1) , 1 1 24

1 0.2ssss

R Re k

k e 则

1)(,2

1)(1)( 2 sHttttr

(2) 因为 0 型系统在速度输入和加速度输入下的稳态误差为无穷大 , 根据叠加原理 ,ess=∞

Page 141: The Principle of Automatic Control 自动控制原理

141

The Principle of Automatic Control 2008

Steady-state error of system with disturbance

•内容:研究输入为 0 ,只有扰动输入的系统稳态误差

•目的:研究系统的抗扰动 (anti-disturbance) 能力以及提高抗扰动的措施

Page 142: The Principle of Automatic Control 自动控制原理

142

The Principle of Automatic Control 2008

理想情况下 , 系统对于任意形式的扰动作用 ,其稳态误差应当为 0,但实际上这是不可能的。

如果输入信号 R(s)=0,仅有扰动 N(s) 作用时 , 系统误差 E(s)=0-C2(s) or e(t)=-c2(t):

0

( )( ) ( )

1 ( ) ( ) ( )

( )lim ( )

1 ( ) ( ) ( )

p

c p

pss

sc p

G sE s N s

G s G s H s

G se s N s

G s G s H s

N(s)R(s)

Gc(s)C(s)E(s)

H(s)

Gp(s)-

与开环传递函数 G(s)=Gc(s)Gp(s)H(s)及扰动信号 N(s) 有关 ,还与扰动作用点的位置有关。

Page 143: The Principle of Automatic Control 自动控制原理

143

The Principle of Automatic Control 2008

r(t)=0

-C(t)

12

Ts

k

)(1)( 0 tMtn

0ks

k1

2

0

00 1 2

( ) :

1lim1

( 1)

0

sss

a

kMTse s

k k k ss Ts

中1 2

0

00 1 2

0 0

( ) :

( 1)lim

1( 1)

/

ss s

b

k kMs Ts

e sk k k s

s Ts

M k

r(t)=0

-

C(t)1

2

Ts

k

)(1)( 0 tMtn

0ks

k1

(a)

(b)

Page 144: The Principle of Automatic Control 自动控制原理

144

The Principle of Automatic Control 2008

在扰动作用点之前的前向通路中增加一个积分环节用( PI regulator )代替

(c)

r(t)=0

-

C(t)1

2

Ts

k

)(1)( 0 tMtn

)1

1(0

0 sTk

s

k1

How to eliminate the error caused by disturbance(如何消除扰动误差 )?

0

)1()

11(1

)1(lim 0

21

00

21

0

s

M

Tsskk

sTk

Tsskk

ses

ss则

Page 145: The Principle of Automatic Control 自动控制原理

145

The Principle of Automatic Control 2008

为了减小扰动作用引起的稳态误差,可以提高扰动作用点之前传递函数中积分环节的个数和增益。而这样都会降低系统的稳定性,而提高开环增益还会使系统动态性能变差,有些控制系统既要求有较高的稳态精度,又要求有良好的动态性能,利用上述方法难以兼顾。为此我们采用按输入补偿或按扰动补偿的复合控制。

Page 146: The Principle of Automatic Control 自动控制原理

146

The Principle of Automatic Control 2008 Example 2:

求图示系统的 essn 。

5s

r=0(0.1s+1)(0.5s+1)

2 c(t)

(1)

n(t)=1(t)

5s

r=0(0.1s+1)(0.5s+1)

2 c(t)

(2)

n(t)=1(t)

解:(1)

C(s)= s(0.1s+1)(0.5s+1)+105 (0.1s+1)(0.5s+1)

s1

∵系统稳定

(2)C(s)=

s(0.1s+1)(0.5s+1)+10

2s s1

∴essn= -limsC(s) = -1/2s→0

∴essn= -limsC(s) = 0s→0

几点说明①增益

② 型别③差异

Page 147: The Principle of Automatic Control 自动控制原理

147

The Principle of Automatic Control 2008

设计 Gr(s)(1) 求图示系统 r(t)= 1(t)+2t 时

输出端定义的误差 ess 。

(2) 为了消除稳态误差 ,试设计 Gr(s),使ess = 0,并绘制出校正后的结构图。

s(s+4)32

0.5

R(s) C(s)

ess = 1

Page 148: The Principle of Automatic Control 自动控制原理

148

The Principle of Automatic Control 2008

Example 3

已知图示系统的调节时间 ts= 0.3秒,

试求 r(t)=3t 时输出端定义的误差终值 ess 。

0.01s1

kh

R(s) C(s) Φ(s)=1/kh

0.01s/kh+1

ts=3T=0.03/kh=0.3

∴kh=0.1

E(s)=1+G(s) H(s)

1H(s)R(s)

ess= 3)10s(s

30

Page 149: The Principle of Automatic Control 自动控制原理

149

The Principle of Automatic Control 2008

设无零点的单位反馈二阶系统 h(t) 曲线如图所示,1、试求出该系统的开环传递函数及参数;2、确定串联校正装置的传递函数,使系统

对阶跃输入的稳态误差为零。

0 1

1.25

0.95

05.0ess 95.0

3.0% s1tp

1s1.4s786.1

19)s(G.1 2

s

k)s(G.2 c

c 12.0k0 c

346.3344.0 n ,=

Page 150: The Principle of Automatic Control 自动控制原理

150

The Principle of Automatic Control 2008

Summary

• 时域分析是通过直接求解系统在典型输入信号作用下的时域响应来分析系统的性能。通常是以系统阶跃响应的超调量、调整时间和稳态误差等性能指标来评价系统性能的优劣。

• 二阶系统在欠阻尼时的响应虽有振荡,但只要阻尼比取值适当(如 0.707 左右),则系统既有响应的快速性,又有过渡过程的平稳性,因而在控制工程中常把二阶系统设计为欠阻尼。

Page 151: The Principle of Automatic Control 自动控制原理

151

The Principle of Automatic Control 2008

• 如果高阶系统中含有一对闭环主导极点,则该系统的瞬态响应就可以近似地用这对主导极点所描述的二阶系统来表征。

• 稳定是系统能正常工作的首要条件。线性定常系统的稳定性是系统的一种固有特性,它仅取决于系统的结构和参数,与外施信号的形式和大小无关。不用求根而能直接判别系统稳定性的方法,称为稳定判据。稳定判据只回答特征方程式的根在 s平面上的分布情况,而不能确定根的具体数值。

Page 152: The Principle of Automatic Control 自动控制原理

152

The Principle of Automatic Control 2008

• 稳态误差是系统控制精度的度量,也是系统的一个重要性能指标。系统的稳态误差既与其结构和参数有关,也与控制信号的形式、大小和作用点有关。

• 系统的稳态精度与动态性能在对系统的类型和开环增益的要求上是相矛盾的。解决这一矛盾的方法,除了在系统中设置校正装置外,还可用前馈补偿的方法来提高系统的稳态精度。

Page 153: The Principle of Automatic Control 自动控制原理

153153

End of Chapter 3End of Chapter 3

Thank you all!Thank you all!