• Nem Talált Eredményt

STABILITY TEST OF LINEAR CONTROL SYSTEMS WITH DEAD TIME COMPENSATED BY PID ACTION CONTROLLER

N/A
N/A
Protected

Academic year: 2022

Ossza meg "STABILITY TEST OF LINEAR CONTROL SYSTEMS WITH DEAD TIME COMPENSATED BY PID ACTION CONTROLLER "

Copied!
14
0
0

Teljes szövegt

(1)

STABILITY TEST OF LINEAR CONTROL SYSTEMS WITH DEAD TIME COMPENSATED BY PID ACTION CONTROLLER

By

IVI. HABERMAYER

Department of Automation, Technical liniYersity, Budapest (Received September 30, 1969)

Presented by Prof. Dr. F. CS_.l.KI

The stability of linear control systems with dead time is rather difficult to determine because of the existence of the exponential function appearing in the system characteristic equation. The use of a digital computer facili- tates, however, to determine the variation of the stability region of a given control system with dead time for various parameter values. These critical loop gain values make it easier to design a given control system.

In a previous paper [1] it 'was investigated how to determine the stability region variation of linear control systems containing second order lag and dead time with a unit feedback compensated, generally speaking, by a PID action controller as shown in Fig. 1. Later graphs were presented giving the

Control/er System

_ exp (-s?;-)

G(sJ- 1+2:S Ts+T2s2 Xo

Fig. 1

critical loop gain Kcr versus dead time with system time constants as param- eters for various damping factor:: values. In these papers the above system was compensated by P, I [10] and PI [11] action controllers, respecth-ely.

The present paper gives the values of KcI' versus dead time in case of a PID controller for different parameter values.

Proportional-plus-integral-plus-derivative control

The transcendental equation determining the critical angular frequency

Wcr belonging to Kcr with the assumption Ti = 1 from (11) of [11] IS:

(1)

(2)

182

50 KeF

t

= 2 - i - - - . . . L

;0

.5

:],2

],1

0,2 0,1

J,02~---+----~·~-

]01~---4---·-

:"!005f---t---

0,05 0,1 0,2

JI. HABERMAYER

0,5

Fig. 2

2 5 10 fC/Tt

(3)

STABILITY TEST OF LINEAR CONTROL SYSTEMS 183 Kcr 10

{= 2 -. ---~----, - - - , -~--~----.--

1,75/1"""--i--_.J...._

5

tJ,9

2 O~, --'--.f=="'l--=i

0,5 0,2

:1 "1

<.11 '

0,2

0,1

0,05

0,02~---4---r---+---

0,05 {),1 0,2 0,5 2 5 10 7:'/l]

Fig. 3

The critical loop gain

Kcr

with arbitrarily accurate approximation values of Wcr may be determined by the following simple algebraic equation:

Wer V(l w~r T2)2+ (2CTwerF Kcr

=

--=--'---'-:;;::;:==;:;:;=~~=7,=---'-

V(l-Td w~r)2+w~r (2)

Figs 2 through 16 show the variation of the stability region as function of the dead time if 0.05

< T/T

i 10 for the following time constant ratios and parameter values:

Td/T

= 0.2, 0.6, 1,

T/T

= 0.2, 0.6, 1, 4, 10,

C =

0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.25, 1.5, 1.75, 2.

6 Periodica Polytechnica El. XlY/2.

(4)

184 JI. HABERJIAYER Kcr 10

5

2

0,5

0,1--;---1-_ _ _

0,2

0,1

7:;'T = 1

0,05 0,1 0,2 0,5 2 5 10 '(;-jT,

Fig. 4

The figures are traced in log-log scale for the sake of clearness. In some figures [11, 16.]

Kcr

curyes belonging to some' parameters are omitted to make the diagrams clearer.

The figures demonstrate that in case of low

Tdlr

values the behaviour of the control system shown in Fig. 1 may be substituted by a eontrol system compensated by a PI action eontroller.

The eorrectness of the latter statement is obvious. By the help of (12) from [1] the critical loop gain using the over-all transfer function of the open loop system is:

(5)

STABILITY TEST OF LLYEAR CO,YTROL SYSTEjIS

Kef 5

0/7: = 0/2

re

2,

1.75,' "

2 1.5 ... "-'- 1,25- 1 1 - 0,9 o'B~--

D,7 . ],0- 0,5 [)5 !J,Lt 0;3 02/ , 0:1//!

0,2

0,05 0,1 0,2 0,5 2 5

Fig. 5

J= s/7:;::: 0,2

0,5

o,2~---~----~---~

0.05 0,1 C,2 0,5 2

Fig. 6

6*

185

J=2 7,5

D5

u:; C,]

(6)

186

KCf 50

20

70

5

2

0,5

0,2

D,1

/ I,

0,9,/ / /1/'

M. HABERJ\L4YER

a~//

a 7/ "

H:----"'.f---"<--'t-i~-\-\:-\-+++t__--_+_--_t_---_j

I ! /

0,6//

I I I

0,5//

o,lf; I

I

0,3/

0,/

To/'"t=0,6

a1-~--~~~~~4.~~~~~YT--~---r--~

7:/1'" 0,2

a~~---~--~~- ...

- - - - + - - - - + - - - I

: --- ···-1 - -~r

i

0,05 0,1 0,2

0,5 2

10 7:/li

Fig. i

(7)

STABILITY TEST OF LINEAR CONTROL SYSTEJIS

KCI' 10

5

2

0,2

0,1

o,02~---~--~----~---l---r---r---r-~~

0,05 0,1 0,2

Mter some arrangements:

0,5

Fig. 8

2 5

187

Kcr

=_ (TJr)(s7:)

.1

+2'(T/7:)(s7:) + (T/7:)2(s7:)2

1

+ (TJ7:)(s7:) +(~/7:)(I;/7:)(S't)2

e-st

(2)

On the basis of equation (3) it can be stated that with decreasing

Td/7:

the critical loop gain more and more approaches the values of Kcr belonging to·

a control system with second order lag and dead time compensated by a PI controller. Relation (3) indicates at the same time that with increasing deadl time the latter behaviour more and more applies.

(8)

188 J'f. HABERJfAYER

:0 I<cr

5 1;2 -

1,75~~

1,5./

0/

7>0,5.

~

1 / 1,25 / '

2 0,9

0,8 0,7 0,5 0,5/

0/1/

0,5

od, '

0,2 0,1

0,2

0.1

7;/T; 1 0,J5

0,05 0,1 0,2 0,5 2 5 10 7:"/TJ

Fig. 9

Fo.r instance, let be: T d!T = 0.2, TIT

=

1,

C

= 0.5, T

=

10. On the basis o.f Fig. 4 o.f [11] the criticallo.o.p gain is: K erp ]

=

0.06. Fro.m Fig. 4 o.f present paper K erPID = 0.061. The difference between KcrPII~=lo and KcrpIDlr=lo is seen to. be rather small.

On the other hand let be: Td/T

=

0.2, TjT = 1, I;

=

0.5,

-c

= 1. The co.rrespo.nding criticallo.o.p gain values are: KcrPIlr~l

=

0.628 and KcrPID!T=l =

= 0.73, respectively. The difference bet'ween the last two. Ker is mo.re signif- icant.

Fro.m the transfer functio.n (3) o.f the o.pen lo.o.p system fo.r very high Td/-C values, i.e. fo.r TdjT -+

= ,

the critical lo.o.p gain is appro.ximatively:

K / ' 8 - _1_ 1+21;(T/T) (sT)+(T/T)2 (S-C)2

er T

d S e-sr

(9)

2

Kcr j=2

1.25

STABILITY TEST OF LINEAR C01\TROL SYSTE.HS

'-~~~~~

OD -- 0.,:; 0.8

0,2

Kcr 2

0,5 0,7 O,f 0,5

DJ

0,2 0,1

0,0.:; 0,

"

0.,2 - - - ---

0,1

0,2 0.,5 2

Fig. 10

0,2 0,5 2

Fig. 11

I

,70/'C=0,61

189

5 10 '011]

5 10 7:iii

Consequently, with increasing

TdlT

the control system may more and more be regarded a system with second order lag and dead time compensated by a PD type controller. For very high values of

TdlT

in comparison with the

TIT

ratio the system may be substituted by a system with dead time com- pensated by a PD action controller.

(10)

190

Kcr 50

20

10

5

2

0,5

0,2

0,1

t;:: 2 _, , 1,75~J 1,5 - 1,25/, 1/ ' /

////1

o,€!/!I/ '~ __

M. HABERMAYER

0,8//1;' :

. ! / I

0, VI-f· I/"+! ----"'~"-~___':_t__t_H_r_t_rtt-'__,_----~--

0.6//;

0,5//

0."1/ ~-

0.3;

0,2 0,1

7:/T = 0.2

a02r---~--- .

001 ,

'---t---I" -'--

, ,

OOO:?t---t---;---,---;----t---j---'~

'_1.'-.,":: 0,5 2 5 10 1,177

Fig. 12

(11)

STABILITY TEST OF LINEAR CONTROL SYSTEMS 19i

Kef 10

5

Tj;/&-= 1 2

0,5

7:/T'" 0,6 0,1 r---r---+----_4---

O,05r---+---+---+---r---I----i-~..._-*~~~

O'02~----_r---_+--_4---+---

0,05 0.2 05 2 5

Fig. 13

(12)

192 .1[, HABERilUYER

<::r 10

5

-

0,9 0,8 0,7 0,6

{)5

/

04 0,3 ,

.~;5 0,2 0,1

1]2

)1

],05

0,05 121 0,2 0,5 2 5 10 '7:/ Tj

Fig. 14

2 ~~~----~~.--~.

Kcr

- [:2

1,75 1.5 - 1,25

0,5 1 .-

0,9 0,8 0,7 0,2 0,6 0,5 0,1

0,1 7:/T = I;

0,05 [,.1 0.2 0,5 2 5 70 7:/T;

Fig. 15

(13)

Kcr 2

0,5

02

r

2

1,75 i,5"

i,25~

1

:J,05

STABILITY TEST OF LINEAR CONTROL SYSTEMS 193

G'/T = 10

DJ .~ D,2 0,5 2

Fig. 16

Conclusion

Knowing the critical loop gain belonging to a given lineal' control system with second order lag and dead time compensated by P, I, PI and PID controller, respectively, one can easily determine what kind of controller proves to be convenient.

In [11], for low dead time values a proportional compensation was found to be preferable. For high dead time values the integral action control proves to be best. For medium dead times the most advantageous type of compensation will be determined by the parameters of the control system.

If the transfer function of the second order lag has the form:

y ( s ) = - - - - -1 (l+s7;.)(l+s7;)

where

Tl

and

T2

are the time constants of the second order lag, Figs. 8 on page 10-12 of [9] shows the preferable kind of compensation. Based on approximate measurements made on an analogue computer, the figure gives the advantageous compensations to be used with a second order lag and dead time system, as function of the dead time and the time constants.

Summary

The present paper gives the stability region variation of the linear continuous control system "With second order lag and dead time compensated by proportional-plus-integral-plus- -derivative PID action controller for different parameter values. The critical loop gain K"

(14)

194 M. HABERMAYER

values were evaluated by a digital computer. TheKcr diagrams for 0.05 ;:;;:; _ITi ;:;;:; 10 are plotted in log-log scale for the sake of clearness ,\ith the system time constants and the damping factor ~ as parameters.

Previous papers investigated the variation of the critical loop gain in cases of P, I [10] and PI [11] compensation.

The Kcr = Kcr(r) diagrams of [10, 11] and the present paper help to choose the most advantageous action controller for different parameter values.

References

1. CS . .\.KI, F.-HABERMAYER, M.: Stability Test of Linear Control Systems with Dead Time.

Periodica Polytechnica, El. Eng. 12, 311-318 (1968).

2. CSAKI, F.: Control Dynamics (In Hung.) Akademiai Kiad6, Budapest 1966.

3. CSAKI, F.-BARS, R-BARKI, K.: Automatika 1. Tankonyvkiad6, Budapest 1966.

4. CS . .\.KI, F.-BARs, R.: Automatika H. Tankonyvkiad6, Budapest 1966.

5. SOLIlIIAN, J. 1.-AL-SHAIKII, A.: A State-space Approach to the Stability of Continuous Systems ,dth Finite Delay. Control, pp. 554-556. Oct. 1965.

6. SOLIMAN, J. 1.-AL-SHAIKH, A.: A State-space Approach to the Stability of Continuous Systems ',ith Finite Delay. Part. H. Control, pp. 626-628 Nov. 1965.

7. CHOKSY, N. H.: Time Lag Systems. Progress in Control Engineering, Volume I, London 1962.

8. EISENBERG, L.: Stability of Linear Systems with Transport Lag. IEEE Transactions on Autom. AC-ll, 247-25'l (1966).

9. GRABBE, I1L-R.um, S.- WOOLDRlDGE, D. E.: Handbook of Automation Computation and Control. Volume 3, 1961.

10. HABERMAYER, M.: Stability Test of Linear Control Systems with Dead Time by Digital Computer. Periodica Polytechnica, El. Eng. 12, 443-452 (1969).

11. HABERMAYER, M.: Stability Test of Linear Control Systems ,\ith Dead Time Compensated by PI Controller. Periodica Polytechnica, El. Eng. 14, 53-59 (1970).

12. ROOTS, W. K.: Electro-heat Control Research at the University of Aston. HIrd Conference on Electro-heat, Budapest, 1969, pp. 621-639.

Maria HABERl\1AYER, Budapest XI, Egry J6zsef u. 16. Hungary

Hivatkozások

KAPCSOLÓDÓ DOKUMENTUMOK

This paper investigates the bounded input bounded output (BIBO) stability in a class of control system of nonlinear difference equations with several time delays.. The proofs are

Similarly to robot control, human motor control involves a reaction delay (dead time), sensory uncertainties (quantization) and saturation at some level of the control

This paper deals with the stability problem of nonlinear delay dif- ference equations with linear parts defined by permutable matrices.. Several criteria for exponential stability

The results of frequency domain analyses of control systems descrihed by linear, second-order lag transfer functions with dead time in the case of serial PID

Procedure in the ALGOL program language for determining the stability region of a linear control system with second order lag and dead time with a unit feedback

The present paper investigates the stability region yariations of linear control system with second order lag and dead time. compensated by P and I typc component:;

should have negative real parts.. With the help of the POl&#34;TRYAGIN method the critical loop gain and thus the region of stability may be determined with- out the

In a following paper [10] we presented graphs showing the stability region variations in function of the dead time and the system time constants for controllers