Automatic Voltage Regulator and Fuzzy Logic Power System Stabilizer
Automatic Voltage Regulator and Fuzzy Logic Power System Stabilizer
Automatic Voltage Regulator and Fuzzy Logic Power System Stabilizer
L.H.Herron
A.Kalam
Abstract
This paper describes the design and implementation of automatic voltage regulator (AVR) and fuzzy logic power system stabilizer (FLPSS) for single machine infinite bus power system. The
AVR was designed using z-domain analytical design method. The proposed
FLPSS employes two nonlinear fuzzy
membership functions to improve its performance. The design and digital simulation studies are carried out using
MATRIXx-a large control system design
and simulation software package. The
design is implemented in a Power System Laboratory with an IBM-486 computer acting as the real time controller.
Both simulation and implementation results show that the proposed PC-based
AVR and FLPSS are very effective.
Keywords: Voltage control, fuzzy logic applications
'ing [1,2].The application of a power system stabilizer (PSS) is to generate a supplementary stabilizing signal, which is applied to the excitation
control loop of a generating unit, t o introduce a
positive damping torque.
In addition to performing the primary control
functions traditionally offered by the analog controls, the PC-based controls have a far greater
deal of flexibility and ability to implement sophisticated control algorithms.
This paper presents a n effective and efficient
Automatic Voltage Regulator (AVR) and Fuzzy
Logic Power System Stabilizer (FLPSS) design
which can be easily implemented by computer facility with high accuracy.
The complete system has been simulated using a
MATRIXx software package on Sun workstation.
The digital AVR and FLPSS are implemented using an on-line 486 PC. Both simulation and implementation results for single machine infinite bus
system show that the proposed PC-based AVR
and FLPSS are very effective.
The system configuration for the single machine
infinite bus system is shown in Figure 1.
Introduction
2
2.1
Design of AVR
Mathematicle model
The full model for single machine infinite bus system is a 7th order model. A machine model chosen for power system dynamic studies depends not
only on the nature of the problem, but also on
293
as :
0.2622
G u ( s )= 1 + 0.2126s
The exciter is modeled using field drive unit and
the time constant of the unit is found to be very
small compared t o the significant time constant of
the system under study and is therefore, neglected.
The transfer function of the field drive unit is:
G F D U (=
~A
)
(3)
where A is the gain of the field drive unit and is
found to be 25 in this case. The three phase line
voltages are transformed t o a proportional DC signal which is measured by the A/D converter. The
sensor circuit, which rectifies, filters and reduces
the terminal voltage t o 5V for comparison, whose
transfer function is found experimentally and is
given by:
O.O29z-'
(9)
1 - O.889~-'
From the given system transfer function the output sequence of the system can be expected to
satisfy the transient criteria. It is required that
the system exhibit a finite settling time with zero
steady-state error.
G ( z )= A *
G,(S) = -
(4)
1 rs
Again the time constant of the sensor circuit T
is neglected as it is very small compared to the
machine time constant. T h e sensor DC gain B
is found to be 0.0042. To design a digital controller the system transfer function must be transformed from s-domain to the z-domain. For a system transfer function G p ( s ) the
, z-domain transfer
function G p ( z ) ,can be obtained by using the following equation:
F ( 2 ) = a,
+
+
+
+
F ( z ) = 3 5 . 2 ~ 3~1 ~. 3 1 8 +
~ 2~ 7~ . 8 1 6 +
~~
2 4 . 7 4 4 ~ - ~2 2 . 0 0 8 +
~ ~1 9~ . 5 6 ~ -+~
1 7 . 4 8 1 5 . 4 5 6+
~ 13.752z-'+
~~
12.216~-lo 1 0 . 8 4 8 ~ - ~+' 9.672~-" (12)
2.2
(10)
(5)
G p ( s )= A
+ q z - 1 + .. .+UNZ-N
+ T&,s
AVR Design
+
+
+
+
B F ( z ) = 0.14672-1
0.1304~-~ 0.1160~-~
0 . 1 0 3 1 ~ - ~0 . 0 9 1 7 ~ - ~0 . 0 8 1 5 ~ - ~0 . 0 7 2 5 ~ - ~ +
0.0644~-~+0.0573z-~+0.0509~-~~+0.0453z~~~+
0.0402~-'~
Thus the controller transfer function can be
rewritten as
G A V R ( Z=
)
5.0586 - 1.2345z-12
*A
1- BF(Z)
(13)
Figures 3 and 4 show the terminal voltage performance corresponding t o different sudden load
294
changes. The solid line shows the terminal voltage performance with the designed AVR. The dotted line shows the terminal voltage without the
AVR. It is obvious that the system equipped with
the designed AVR kept the output constant under
disturbance conditions.
3.2
FLPSS Design
Us(t) = U S ( k ) ,
-*
2
3
3.1
(14)
The origin in the phase plane is the desired equilibrium point. All the control effort should be directed to moving the current condition p ( k ) towards the origin as quickly as possible.
The accelerating control of the study unit is
achieved by applying a negative stabilizing signal
to the excitation loop, as the electrical output of
the study unit can be decreased by the negative
stabilizing signal. Correspondingly, decelerating
control is achieved by applying a positive stabilizing signal to the excitation loop with the increased
electrical output through the positive stabilizing
signal.
Two fuzzy nonlinear membership functions,
N { B i ( k ) } and P { & ( k ) } ,are defined for the proposed FLPSS as shown in Figure 5 to represent
both the sectors A and B respectively. The term
Bi(k) indicates the phase angle of the point p i ( k ) .
By using these membership functions, the stabilizing signal is computed as follows:
VJ(k) =
295
G( k ) [ N{ Bi (IC)}
for D( k ) >_ D, ,
G ( k )= 1.0,
(20)
D(k) = I P(k) I
(21)
The term G ( k ) indicates the gain factor at the
time t = kAT, and G ( k ) is given by a nonlinear
function.
The maximum value of the stabilizing signal
U,,
depends on the generating unit. Distance
parameter D, and angles a and p can be adjusted
t o their optimal values according to a performance
index which is defined as follows:
In order to verify the design for the fuzzy logic controller as well as the simulation results, the fuzzy
controller was implemented and tested on a onemachine infinite bus system.
After the system is run up to speed, the generator was synchronized and connected to an infinite
bus.
In all the tests, the steady state was defined
with an electrical power of 1625W a t 0.9 power
factor. The tests were examined after a sudden
load change. Different load has been applied t o
test the robustness of the system response.
Figures 8-13 show the rotor angle and speed
deviation corresponding to different sudden load
changes respectively. Figures 8,lO and 12 show the
system response without PSS. Figures 9,11 and 13
are the results with a FLPSS. To verify the robustness of the system performance, different operating loading conditions have been applied and
similar result has been obtained. Figure 12 and 13
are the results under another operating condition.
The study shows that the system with a FLPSS
increase the system damping dramatically.
J =
[tkA~(k)]
(22)
k=l
6
for Bo
0
(23)
+ 180 + a 5 8i 5 360
Conclusions
Implementation Study
Simulation Study
296
References
LARSEN,E.V.,SWANN,D.A.:Applying
Power System Stabilisers.IEEE Trans.,PAS100, 1981, p p . 3017-3046.
AND E RS 0N ,P.M .,FO UAD ,A.A., : Pow e r
System Control and Stability. Iowa State
University Press, Ames, Iowa, 1977.
DEMELLO,F.P.,CONCORDIA,C.:
Concepts
of Synchronous Machine Stability as Affected
by Excitation Control,IEEE Trans., PAS88,1969, p p . 316-329.
/I\,,
/I\,,
95tlO
180
275305
350
APPENDIX
One-machine infinite bus system generator unit
data:
x d = 1.027pu
x; = 0 . 4 7 9 ~ ~
xq = 0 . 4 8 9 ~ ~
Ti, = 0.345s
H = 0.764s
W b = 314rad/sec
R, = 0 . 0 2 ~ ~
x, = 0.5pu
Lwm-
297
-10
-20
-25
-301
0
'
'
'
'
'
IO
'
12
'
14
'
16
I8
20
tlmc m sccmda
Figure 11: Speed deviation corresponding to sudden load change (with FLPSS)
-=I
-30
'
'
'
'
'
IO
12
14
16
18
20
-15Li
U
,
-200
'
'
IO
'
I2
'
14
'
16
'
18
8
I
IO
I2
I
14
I
16
I
18
4'
0
'
20
tunc m sccmds
Figure 10: Speed deviation corresponding to sudden load change (without PSS)
298