International Journal of Science and Engineering Applications

Volume 8–Issue 07,180-185, 2019, ISSN:-2319–7560

Performance Analysis of Automatic Voltage Regulator in

Power Generation System
Wint Yu Yu Zaw
Electrical Power Engineering
Department
West Yangon Technological
University
Yangon, Myanmar

Abstract: The system voltage is one of the most important parameters, which determines the power quality. The stability of the
system voltage is critical for the power system. This paper investigates the analysis of the automatic voltage regulator (AVR) system
controlled by stabilizer and PID controller. As the work of the AVR is to maintain the synchronous generator terminal voltage. AVR
controller simulation model created in power system based on real data from Baluchaung No.2 Hydro Power Plant (Myanmar) to show
that the proposed method is effective. The system modeled consists of the amplifier, exciter, sensor and stabilizer or proportional-
integral-derivative (PID) controller. Their characteristics and behavior are also briefly discussed. The transfer function controller
blocks use linear block model and control techniques in MATLAB/ Simulink environment. By examining the parameters of the
stabilizer and PID controller, the generator terminal voltage variations are predicted. Simulation results demonstrate that the proposed
controllers provide better performance characteristics with respect to settling time, oscillations and overshoot.
Keywords: Automatic Voltage Regulator (AVR), electric power system, voltage response, Stabilizer, PID controller

1. INTRODUCTION 2. AUTOMATIC VOLTAGE

The system that provides electrical energy flow from REGULATOR (AVR)
electricity generation centers to consumption centers is called The rules of AVR are to maintain the generator terminal
an interconnected system. This energy flow should be voltage magnitudes at the specified level. AVR system
continuously provided on demand for system stability. The consists of four main components there are amplifier, exciter,
active and/or reactive power change occurring in demand generator and sensors. The mathematical model and transfer
shows itself in two different ways in the electricity generation functions can be to linear taking into account the time
center. The first one is the case resulted in the change of constant and ignoring saturation or other nonlinear. In Figure
frequency of the system, that is, rotation speed of the (1), the schematic diagram of an AVR system is presented.
generator. Load frequency control systems are used to hold at
the desired frequency. The second one is the change occurring
in the terminal voltage of the generator due to the change
especially in reactive power. AVR is the control mechanism
that detects variations in the output voltage of the generator
and provides a controlling excitation current. However, AVR
system without any controller will provide slow responses and
may cause instability. Thus, need to provide proper control of
generator by AVR system to ensure generators operated at
safe and stable conditions for normal operations.
Control of AVR system has been proposed in
BaluchaungNo.2 Hydro Power Plant (Myanmar).The author
designed an AVR system using MATLAB/Simulink and the
performance of the system was evaluated using PID controller
and stabilizer. In those studies, systems using stabilizer with Figure 1. Schematic diagram of an AVR system
and without time-delay has been investigated. We also
analyzed the mentioned system after introducing the
automatic voltage regulators (AVR) and PID controller into 3. MATHEMATICAL MODELLING OF
the system. In this work, the combined action of the two AVR SYSTEM
controllers will be tested as a single controller to determine For control system performance, mathematical model of each
their resultant effect instead of using them separately in the component is needed. There are two methods to
AVR system. It was deduced that the proportional controller mathematically model an element one is transfer function
(KP) have the effect of reducing rise time, the integral method and other is state variable method. State variable
controller (KI) have the effect of eliminating steady - state approach is preferable for both non-linear as well as linear
error, and the derivative controller (KD) have the effect of systems. The transfer function approach is simpler and
increasing stability of the system. applicable for linear system only.
For transfer function approach, some assumptions and
approximations are taken to linearize the non-linear model.
Variation in the magnitude of system voltage depends on
transient reactance of the synchronous machine while the

www.ijsea.com 180
 International Journal of Science and Engineering Applications
Volume 8–Issue 07,180-185, 2019, ISSN:-2319–7560

duration of variation is decided by the generator time relatively inefficient, particularly if the loops have response
constant. Being a static device, AVR offers minimum time times on the order of minutes or longer.
delay and brings the system voltage to its nominal value
quickly. The terminal voltage is sensed by AVR and is The choice of method will depend largely on whether or not
maintained at its rated value by regulating the excitation. the loop can be taken "offline" for tuning, and on the response
time of the system. If the system can be taken offline, the best
3.1 Amplifier Model tuning method often involves subjecting the system to a step
Transfer function of an amplifier is modelled by a gain change in input, measuring the output as a function of time,
and using this response to determine the control parameters.
and a time constant given by:
Ka Transfer function of a PID controller is modelled by a gain
TFA  and a time constant given by:
1  s a Eq: 1
where, Ka and τa represent the gain and time constant of KI
TFPID  K P   KDs
the amplifier. The values of Ka are in the range of 10 to s Eq: 6
40 and τa are in the range of 0.02 to 0.1 sec.
where, KP, KI and KD represent the gain constants of PID
3.2 Exciter Model controller.
Transfer function of an exciter is modelled by a gain and a Table 1. Typical values for parameters
time constant given by:

Ke
TFE 
1  s e Eq: 2
where, Ke and τe represent the gain and time constant of the
exciter. The values of Ke are in the range of 1 to 2 and τe are
in the range of 0.4 to 1.0 sec.

3.3 Generator Model

Transfer function of a generator is modelled by a gain and a The excitation system model of AVR system is shown in
time constant given by: Figure (2).

Kg D Vref (s) + D Ve (s) Ka D Vr (s) K e D Vf (s) K g D Vt (s)

TFG  1   as 1   es 1 gs
1  s g -
Amplifier Exciter Generator
Eq: 3 D Vs (s)
K s
where, Kg and τg represent the gain and time constant of the 1   ss
generator and load are dependent. The values of Kg are in the Sensor
range of 0.7 to 1.0 and τg are in the range of 1.0 to 2.0 sec
from full load to no load. Figure 2. Block Diagram of AVR System

3.4 Sensor Model

Transfer function of a sensor is modelled by a gain and a time D Vref (s) + D Ve (s) Ka D Vr (s) K e D Vf (s) K g D Vt (s)
1  as 1   es 1 gs
constant given by: - -
Amplifier Exciter Generator

Ks
TFS  K f

1  s s
1   f s

Eq: 4 D Vs (s)
Stabilizer

where, Ks and τs represent the gain and time constant of the

K s
1   ss
sensor. The value of Ks is 0.02 to 1.0 and τs are 0.01 to 2.2 Sensor
sec.
Figure 3. Transfer Function Model of AVR System with
3.5 Stabilizer Model Stabilizer
Transfer function of a stabilizer is modelled by a gain and a
time constant given by:
D Vref (s) KI Ka D Ve (s) Ka D Vf (s) Ka D Vt (s)
+ KP   KDs 1 as 1 as 1 as
Kf - s

TF f  PID Amplifier Exciter Generator

1  s f
Eq: 5
D Vs (s)
where, Kf and τf represent the gain and time constant of the
sensor. The value of Kf is 1.0 and τf is 0.01 to 0.06 sec. Ks
1 ss
Sensor

3.6 PID Controller Model Figure 4. Transfer Function Model of AVR System with PID
There are several methods for tuning a PID loop. The most
Controller
effective methods generally involve the development of some
form of process model, choosing P, I, and D based on the
dynamic model parameters. Manual tuning methods can be

www.ijsea.com 181
 International Journal of Science and Engineering Applications
Volume 8–Issue 07,180-185, 2019, ISSN:-2319–7560

D Vref (s) KI Ka D Ve (s) Ka D Vf (s) Ka D Vt (s)

+ KP   KDs 1 as 1 as 1 as
- s
PID Amplifier Exciter Generator

K f

1 fs
D Vs (s) Stabilizer

Ks
1   ss
Sensor

Figure 5. Transfer Function Model of AVR System with

Stabilizer and PID Controller

4. Location of Research
This research was conducted at the Baluchaung No.2
Hydropower Plant. The Baluchaung No. 2 Power Stations
have a generation capacity of (28x6) MW, respectively. Figure 8. AVR Structure of Baluchaung No.2 Hydropower
Electric power generated at this plant is sent to the Yangon Plant
and the Mandalay region; thus Baluchang No. 2 Power Station
plays a significant role in the supply of electric energy in 5. RESEARCH INSTRUMENT
Myanmar. The instrument used in this study is the hardware and
software. Hardware includes a set of computers that are
compatible with the software used, while the software
includes Matlab Sim link.

6. PARAMETERS FOR AVR MODEL

For control system stability, the Routh-Hurwitz array was
constructed and chooses the amplifier gain constant Ka.
Assume AVR parameters as follow:
Amplifier gain constant, Ka -
Amplifier time constant, a 0.05

Exciter gain constant, Ke 1

Exciter time constant, e 0.4

Generator gain constant, Kg 1

Generator time constant, g 0.5
Figure 6. Site Map Sensor gain constant, Ks 1
Sensor time constant, s 0.01

D Vref (s) + D Ve (s) Ka D Vr (s) 1 D Vf (s) 1 D Vt (s)

1  0.05s 1  0 .4 s 1  0 .5 s
-
Amplifier Exciter Generator
D Vs (s)
1
1  0 .01 s
Sensor

Figure.9 AVR Block Diagram

Open-loop transfer function of AVR is

K a (1)(0.1)
Figure 7. Location of Baluchaung No.2 Hydropower Plant G (s) H (s) 
(1  0.05s )(1  0.4 s )(1  0.5s )(1  0.01s )
10000K a

s  124.5s  2545s2  9600s  10000
4 3

The characteristic equation is

10000Ka
1  G(s)H(s)  1  0
s  124.5s  2545s2  9600s  10000
4 3

 s 4  1245s3  2545s2  9600s  10000  10000Ka  0

www.ijsea.com 182
 International Journal of Science and Engineering Applications
Volume 8–Issue 07,180-185, 2019, ISSN:-2319–7560

The Routh-Hurwitz array for this polynomial is 6.1 AVR with Stabilizer
s4
1 2545 10000  10000K a Assume stabilizer parameters as follow:
s3 124.5 9600 0 Stabilizer gain constant, Ks 0.4
s 2
2467.89 10000  10000K a 0
Stabilizer time constant, τs 0.04
s1 504.48K a  9095.52 0
s 0 10000  10000K a
D Vref (s) + D Ve (s) 11 D Vr (s) 1 D Vf (s) 1 D Vt (s)
From the s1 row, to find Ka for control system stability, - - 1  0.05s 1  0 .4 s 1  0 .5 s
Amplifier Exciter Generator
504.48Ka – 9095.52< 0
0 .4 s
Ka< 18.03 1  0 .04 s
Stabilizer
D Vs (s)
For Ka = 18.03, the auxiliary equation from the s2 row, 1
1  0 .01 s
2467.89s2 + 10000 + 10000 (18.03) = 0 Sensor

s = ±j8.78
Figure.11 AVR with Stabilizer Block Diagram
Since, for Ka = 18.03, a pair of conjugate poles on the j axis,
and the control system is marginally stable.
6.2 AVR with PID Controller
PID controller transfer function is
Closed-loop transfer function of AVR is
KI
G ( s) V ( s) GC (s)  K P   KDs
 t s Eq: 9
1  G ( s ) H ( s ) Vref ( s )
Eq: 7
For control stability, the constants KP, KI and KD should
0.01K a ( s  100) choose by tuning, on the response time of the system. The
 4 best tuning method involves subjecting the system to a step
s  124.5s  2545s 2  9600s  10000  10000K a
3
change in input, measuring the output as a function of time,
and using this response to determine the control parameters.
For the system to be stability, the amplifier gain of Ka should
be chosen less than 18.03. Assume PID controller parameters as follow
Figure 10 shows the comparison of effects of Ka for AVR Proportional gain constant, KP 1.0
system.
Integral gain constant, KI 0.25
Derivative gain constant, KD 0.28
D Vref (s) 0.25 11 D Ve (s) 1 D Vf (s) 1 D Vt (s)
+ 1.0   0.28s
- s 1  0 .05 s 1  0 .4 s 1  0 .5 s
PID Amplifier Exciter Generator

D Vs (s) 1
1  0 .01s
Sensor

Figure 12. AVR with PID Controller Block Diagram

6.3 AVR with Stabilizer and PID

Figure.10 Effects of Ka for AVR System Controller
From above Figure, choose Ka= 11, the steady-state response D Vref (s) 0.25 11 D Ve (s) 1 D Vf (s) 1 D Vt (s)
+ 1.0   0.28s
- s 1  0 .05 s 1  0 .4 s 1  0 .5 s
is PID Amplifier Exciter Generator

0 .4 s
1  0 .04 s
D Vs (s) Stabilizer

Eq: 8 1
1  0 .1s
Sensor
11
Vt , ss   0.9167
1  11 Figure 13. AVR with Stabilizer and PID Controller Block
Diagram
The steady-state error is
7. SIMULINK MODEL AND RESULTS
Ve,ss= 1.0- 0.9167 = 0.0833 OF AVR SYSTEM
Figure 14(a) shows an AVR system model in MATLAB
In order to reduce the steady-state error, the amplifier gain Simulink, where a combined simulation block model was
must be increased, which results in an unstable control constructed for the amplifier, exciter, and generator cases in
order to obtain their respective step response.
system. Therefore, choose Ka= 11.

www.ijsea.com 183
 International Journal of Science and Engineering Applications
Volume 8–Issue 07,180-185, 2019, ISSN:-2319–7560

and PID controller cases in order to obtain their respective

step response.

Figure.14 (a) Simulink Model of Simple AVR System

Figure.16 (a) Simulink Model of AVR System with PID

Controller

Figure.14 (b) Terminal Voltage Response of AVR System

Figure.16(b) Terminal Voltage Response of AVR System

From Figure 14(b), the result shows that for an amplifier gain with PID Controller
Ka = 11, the response is highly oscillatory, with a large From Figure 16(b), a proportional gain KP = 1.0, an integral
71.552% overshoot and a long settling time with 8.177 sec. gain of KI = 0.25 and a derivative gain of KD = 0.28 is tuned
Since the steady-state error is 8.33%, there is high steady-state and it is found to be satisfactory. PID controller produced
error and unsatisfactory transient response. about 0.385 % overshoot and settling time of 4.744 sec. The
Figure 15(a) shows an AVR system model with stabilizer in PID controller reduces the steady-state error to zero. It was
MATLAB Simulink, where a combined simulation block observed that the AVR shows 1 p.u. output voltage with PID
model was constructed for AVR and stabilizer cases in order controller. It also shows that low frequency and less
to obtain their respective step response. oscillation.
Figure 17 (a) shows an AVR system model with PID
controller and stabilizer in MATLAB Simulink, where a
combined simulation block model was constructed for AVR,
stabilizer and PID controller cases in order to obtain their
respective step response.

Figure.15 (a) Simulink Model of AVR System with Stabilizer

Figure.17 (a) Simulink Model of AVR System with Stabilizer

and PID Controller

Figure.15 (b) Terminal Voltage Response of AVR System

with Stabilizer
From Figure 15(b), the result shows that a very satisfactory
transient response with steady state error of 5.4% and a
settling time of approximately 2.255 seconds.
Figure 16(a) shows an AVR system model with PID Figure.17 (b) Terminal Voltage Response of AVR System
controller without stabilizer in MATLAB Simulink, where a with Stabilizer and PID Controller
combined simulation block model was constructed for AVR

www.ijsea.com 184
 International Journal of Science and Engineering Applications
Volume 8–Issue 07,180-185, 2019, ISSN:-2319–7560

From Figure 17(b), a proportional gain KP = 1.0, an integral Table 2 shows the comparison results for the AVR system
gain of KI = 0.25 and a derivative gain of KD = 0.28 is tuned with and without controllers.
and it is found to be satisfactory. The response settles in about
3 seconds with no overshoot. The combined controllers Table 2. Comparison results for the AVR system with and
reduce the steady-state error to zero. without controllers

Figure 18(a) shows an overall AVR system model with AVR, Rise Peak Settling Steady
Overshoot
stabilizer, and PID controller in MATLAB Simulink. System Time Time Time State
(%)
(ms) (s) (s) (p.u.)

AVR Closed Loop 71.552 161.49 0.484 8.177 0.92

AVR with Stabilizer 9.341 172.428 0.357 2.255 0.67

AVR with PID 0.458 635.164 4.312 4.744 1

AVR with Stabilizer

1.531 775.096 2.492 2.748 0.72
and PID Controller

The AVR was simulated with stabilizer and PID controller. It

was observed that the AVR shows low frequency and less
oscillation. Figure 18(b) shows the comparison results of the
system. From these results, AVR with PID controller shows
that the reliability of generating system is determined by its
ability for maintaining voltage and frequency within
permissible limit. It can be concluded that the generating
system for AVR with PID controller shows satisfactory
performance.

8. CONCLUSION
In this paper, application of PID controller and stabilizer
inAVR control system with variation in its time constant
andgain was examined. By using PID controller and stabilizer,
Figure.18 (a) Simulink Model of Overall AVR System with notonly the system will have an appropriate steady-state error
Controllers butalso provides good dynamic response with a less overshoot
forthe voltage terminal. Finally, the comparison results of
AVR system were simulatedand effects of some parameters
were investigated.From these results performance response of
the AVR system in generator with PID controller gives
satisfactory settling time and less oscillation. Thus AVR with
PID controller is very suitable for controlling the voltage of
power generation system.

9. REFERENCES
[1] Cohn, N., Control of Generation and Power Flow on
Interconnected Systems, Wiley, New York, 1971.
[2] KavitaGoswami, “Load Frequency and Voltage Control
Figure.18 (b) Terminal Voltage Response of Overall AVR of Two Area Interconnected Power System using PID
System with Controllers Controller”, International Journal on Emerging
Technologies (Special Issue NCETST-2017) 8(1): 722-
726(2017)
[3] Orosun. M.M, “Modeling and Simulation of Automatic
Generation Control System for Synchronous Generator
with Model Predictive Controller”, Zimbabwe Journal of
Science & Technology pp 142 - 157
[4] Power System Analysis, HadiSaadat, McGraw-Hill,
International Editions 1999.
[5] P. Girish1, “Matlab design of AVR for single Area of
Power System”, Journal of Chemical and Pharmaceutical
SciencesISSN: 0974-2115
[6] K.Eswaramma1, Dr.G.Surya Kalyan2, “An Automatic
Voltage Regulator(AVR) System Control using a P-I-DD
Controller”, International Journal of Advance
Engineering and Research, Development, e-ISSN (O):
2348-4470, p-ISSN (P): 2348-6406

www.ijsea.com 185