A Digital-Based Optimal AVR Design of Synchronous Generator Exciter Using LQR Technique

Al-Khwarizmi
Engineering
Journal
Al-Khwarizmi Engineering Journal, Vol. 7, No. 1, PP 82 - 94 (2011)

A Digital-Based Optimal AVR Design of Synchronous Generator
Exciter Using LQR Technique
Ibraheem Kasim Ibraheem
Department of Computer Engineering/ College of Engineering/ University of Baghdad
Email: [email protected]
(Received 18 April 2010; Accepted 23 December 2010)
Abstract
In this paper a new structure for the AVR of the power system exciter is proposed and designed using digital-based
LQR. With two weighting matrices R and Q , this method produces an optimal regulator that is used to generate the
feedback control law. These matrices are called state and control weighting matrices and are used to balance between
the relative importance of the input and the states in the cost function that is being optimized. A sample power system
composed of single machine connected to an infinite- bus bar (SMIB) with both a conventional and a proposed Digital
AVR (DAVR) is simulated. Evaluation results show that the DAVR damps well the oscillations of the terminal voltage
and presents a faster response than that of the conventional AVR.
Keywords: Exciter design, AVR, LQR, state-feedback, SMIB.
1. Introduction
Excitation control of generators is a very
important topic in the field of power systems. A
good excitation control, indeed, has proven to be
very efficient to support the voltage on the power
system, to enhance its transient stability and to
damp its oscillations [1].
The basic requirement is that the excitation
system supply and automatically adjust the field
current of the synchronous generators to maintain
the terminal voltage as the output varies within
the continuous capability of the generator
(Generator considerations), while the excitation
system should contribute to effective control of
voltage and enhancement of system stability
(Power system considerations) [2].
With the emergence of digital technology,
digital based excitation systems are being evolved
and this development has added to the AC and ST
types of excitation control systems [3]. Recent
trend in modern excitation control systems is
towards the use of the digital electronics to
perform control and protection functions [4].
Many features are available in digital
excitation in comparison to analog excitation
systems: greater precision, immunity to
component variations, reduced cost, reduced
hardware and hence less maintenance cost,
availability of on board test features, implement
ability of fuzzy and adaptive or other control
schemes. Microprocessors or microcontroller or
PCs used in these systems give capability to
provide much more information and greater
control. Flexible software or using different
software with the same unit provides greater
choice, possibility of more protection features,
greater memory and self monitoring of PC cards
[4]. In addition, this type of solution is much
cheaper than heavy equipments like FACTS
(Flexible Alternating Current Transmission
System) which are necessary only in very specific
cases, as far as oscillation damping is concerned.
Digital systems are the state of the art systems,
both for new synchronous machines as well as for
modernizing existing older plants. Now with
digital control, the excitation system models are
modified to include proportional gain
P
K and
integral gain
I
K which are adjustable [4].
Studies on the detailed models of digital
excitation systems with SMIB have not been
This page was created using Nitro PDF trial software.
To purchase, go to http://www.nitropdf.com/
Ibraheem Kasim Ibraheem Al-Khwarizmi Engineering Journal, Vol. 7, No. 1, PP 82 - 94 (2011)
83
reported well in the literature and, hence, they
have drawn much attention. Many attempts have
been done in the last years regarding the design of
excitation control systems. The H control
theory has been exploited in [5] to design an
excitation control system using LMI (Linear
Matrix Inequality) to damp power system
oscillations and control of terminal voltage
simultaneously; weighting filters have to be
designed for this technique. The authors in [6]
designed a neurocontroller to replace both the
conventional automatic voltage regulator (AVR)
and the turbine governor of the power systems,
while in [7] a separate neurocontroller for the each
of the AVR and the turbine governor has been
designed. In [8] researchers present a step-by-
step design methodology of an adaptive neuro-
fuzzy inference system (ANFIS) based AVR and
power system stabilizer (PSS) and also
demonstrated its performance in a single-
machine-infinite-bus and a multi-machine power
system through digital simulation. In [9], an
adaptive design of an automatic voltage regulator
(AVR) control scheme for synchronous generators
is introduced in the presence of unknown
variations of the power system operating
conditions; the AVR design is based on pole-
assignment technique and the estimation is
performed by kalman filter. While in [10], an
efficient and powerful design method suitable for
calculating optimal proportional-integral-
derivative (PID) controllers for AVR systems is
proposed. The method is an improved version of
the Discrete Action Reinforcement Learning
Automata (DARLA). In [11], authors employed a
loop-shaping method, which is achieved via a
parameter space approach accomplished by
symbolic method called quantifier elimination
(QE) and apply it to design PI control type AVR
of the excitation control. In [12], a step-by-step
coordinated design procedure for power system
stabilizers (PSSs) and automatic voltage
regulators (AVRs) in a strongly coupled system is
described.
Figure 1 shows a typical continuous feedback
control system. Almost all of the continuous
controllers can be built using analog electronics
[13, 14].
The continuous controller, enclosed in the
dashed square, can be replaced by a digital
controller, as shown in figure 2 below, that
performs the same control task as the continuous
controller. The basic difference between these two
controllers is that the digital system operates on
discrete signals (or samples of the sensed signal)
rather than on continuous signals [13,14].
Fig. 2. Closed-loop Control System.
Fig. 1. Analog Closed-Loop Control System.
K(s)
-

Plant
r(s) y(s) e(s) u(s)
Continuous Controller
D/A and
hold
-
Plant
(t)

Clock
Difference
Equation
A/D
A/D
r(t) r(k) e(k) u(k) (t)
(t)
+
) (s G
) (z K
y(k)
) (z G
zoh
84
K(z)
-
) (z G
zoh

r(k
)
y(k
)
e(k
)
u(k)
In the above schematic drawing of the digital
control system, it can be seen that the digital
control system contains both discrete and
continuous portions.
The clock connected to the D/A and A/D
converters supplies a pulse every
s
T second and
each D/A and A/D sends a signal only when the
pulse arrives. The purpose of having this pulse is
to require that ) (z G
zoh
have only samples of
) (k u to work on and produce only samples of
output ) (k y ; thus, ) (z G
zoh
can be realized as a
discrete function [14].
Now the schematic will be drawn by placing
) (z G
zoh
in place of the continuous portion. By
placing ) (z G
zoh
, a digital control system ) (z K
can be designed which deals only with discrete
functions, see figure 3. When designing a digital
control system, the discrete equivalent (discrete-
state-space or transfer function system) of the
continuous portion must be found so that only
discrete functions are dealt with in this work [14].
That will be done in the next section.
Fig. 3. Closed-Loop Digital Control System.
This paper is organized as follows: a brief
introduction to the digital design of linear
quadratic regulator is given in section 2; section 3
describes the state-space modeling of the power
system excitation control. Evaluation results are
included in section 4. Finally conclusions are
given in section 5.
2. Digital Control System Design with
Optimal Linear Quadratic Regulator
Given an analog linear system with state-space
equation:
) ( u D ) ( x C ) (
) ( u B ) ( A x ) (
t t t y
t t t x
+ =
+ =
-
(1)
where ) (t x is an n 1 real vector, ) (t u is an
m 1 real vector, and ) (t y is an l 1 real vector,
the matrices in the equations are:
A = n n state matrix.
B = n m input or control matrix.
C = l n output matrix.
D = l m direct transmission matrix.
With piecewise constant inputs over a given
sampling period (
s
T ), the system state variables at
the end of each period can be related by a
difference equation. It is obtained by examining
the solution of the analog state (eq. (1)) over
s
T
[15,16].
The discrete time state equation is given by
[15,16]:
) ( ) ( ) 1 ( k u B k x A k x
d d
+ = + (2)
Where
s
AT
d
e A = (3)
}
=
s
T
0
A
d
d B e B
(4)
The output equation evaluated at time
s
kT is
[15,16]:
) ( ) ( ) ( k u D k x C k y + = (5)
The discrete-time state-space representation is
given by eq(2) and eq(5). It should be noted that
the matrices
d d
D C & of the discrete system are
equal to its corresponding matrices of the
continuous system.
One of the difficulties in designing the pole
placement controllers is the selection of closed-
loop pole locations, especially in a large
dimensional state space system. An alternative
approach is to use an optimization based strategy
to design the controller. Minimization of the
square of an error for a linear system results in a
Linear Quadratic Regulator abbreviated as LQR
[16, 17]. The schematic of a full-state feedback
system is drawn in figure 4.
The next step is to assume that all states are
measurable, and find the control matrix (K). The
LQR method will be used to find the control
matrix (K).
85
Fig. 4. Quadratic Optimal Control System.
The LQR design approach involves the
minimization of a weighted sum of quadratic
functions of the state and the control effort. If a
quadratic function of only the states is minimized,
the control effort could become unbounded. To
overcome these difficulties, the quadratic function
of the control effort is minimized as well [15].
Consider the discrete-time state-space system:
) ( ) ( ) 1 ( k u B k x A k x
d d
+ = +

(6)
) ( ) ( ) ( k u D k x C k y + =
(7)
A control law ) ( ) ( k x K k u - = has to be found
so that a cost function [15,16]:
)] ( ) ( ) ( ) ( [ k u R k u k x Q k x
2
1
J
T
0
T
+ =

(8)
is minimized, where Q and R are constant
symmetric state and control weighting matrices to
be selected by the designer, with Q positive
semidefinite and R positive definite. When the
objective function is a sum of an infinite number
of terms, known as the infinite horizon problem,
the steady state solution is the optimal solution
[16, 18].
As some of the states may be allowed to be
zero and all control efforts have a cost associated
with them, some conditions must be satisfied
[16,18]:
1. x x Q x
T
> 0 . (9)
2. 0 x u R u
T
= > 0 . (10)
3. The (
d d
B A , ) is controllable (stablizable).
4. The pair
(
C A
d
, ) is observable (detectable).
The state weighting matrix can be decomposed
as follows [16]:
C C Q
T
= (11)
where C is the output matrix and is positive
definite for Q positive definite and positive
semidefinite for Q positive semidefinite.
Conditions (3) and (4) guarantee the Riccati
equation does not diverge and a steady-state
condition is reached. The resulting algebraic
Riccate equation is in the form [16, 18, 19]:
Q A S B R SB B SB S A S
d
T
d d
T
d d
T
d
+ + =

] ) ( [
1
...(12)
It is called Discrete Algebraic Riccate
Equation (DARE), and it has a unique positive
definite solution. However, the equation is clearly
difficult to solve in general and is typically solved
numerically [19].
The optimal state feedback control law
corresponding to the steady-state regulator is
[15,16, 18, 20]:
) ( ) ( k x -K k u = (13)
SA B SB B R K
T
d d
T
d
1
] [

+ = (14)
Notice that this optimal control law K can be
calculated once and for all at the very beginning.
Unlike other design methods, the full-state
feedback system does not compare the output to
the reference; instead, it compares all states
multiplied by the control matrix ( x K- ) to the
reference (see Figure.4). Thus, it is not expected
to see the output equals the input. To obtain the
desired output, the reference input should be
scaled so that the output equals the reference. This
can be easily done by introducing a feedforward
scaling factor called Nbar. The basic schematic
with Nbar is shown in figure 5 [14, 20].
r(k) = 0
x
y(k)
K
u(k)
) 1 ( u ) 1 ( x C ) 1 (
) ( u ) 1 ( ) 1 ( x A ) (
+ =
+ =
k D k k y
k k B k k x
d d

) 1 ( u ) 1 ( x C ) 1 (
) ( u ) 1 ( ) 1 ( x A ) (
+ =
+ =
k D k k y
k k B k k x
d d

y(k)
K
u(k)
Nbar
r(k)
86
Fig. 5. Quadratic Optimal Control System.
The scaling factor Nbar can be calculated as
follows [15,20]:
x u
N K N N + = (15)
Where

(

=
(

I C
B I A
N
N
d d
u
x
0
.
0
1
(16)
Finally, it is important to present general
guidelines for choosing the state and control
weighting matrices Q and R [15,20]:
1. Both of them must be symmetric with Q that is
positive semidefinite and R that is positive
definite.
2. R is usually chosen to be purely diagonal
matrix while Q is given by eq (11).
3. It is a normal practice to select initial values
for Q and R, carry out the design of the
feedback gain matrix K, and then simulate the
resulting system and view the response. Values
of Q and R are changed as required by the
designer and the procedure is repeated.
3. Modeling of the Power Plant Exciter
System.
The excitation control system is one of the
important factors in the transient study of power
system analysis. It controls the generated EMF of
the generator and therefore controls not only the
output voltage but the power factor and the
current magnitude as well. A typical relationship
between the excitation control system and the
generator is illustrated in figure 6 [21].
Fig. 6. Typical Arrangement of Excitation Components.
The exciter provides a dc power to the
synchronous machine field winding, constituting
the power stage of the excitation system. While
the regulator process and amplifies input control
signals to a level and forms appropriate control of
the exciter. The other protective and limiting
functions are achieved by auxiliary control [2].
The power system considered in this study is
modeled as a synchronous generator connected to
an infinite bus bar. A static excitation system and
automatic voltage regulator (AVR) are employed
to maintain the terminal voltage profile as shown
in figure 7 [22,23].
Fig. 7. Power System Configuration.
Exciter &
AVR
Control input
Transmission line
E
t
Governor
Turbine
Generator
Generator
Auxiliary
Control
Exciter
Voltage
Regulator
Input torque
from
prime mover
Exciter power
source
Desired
Voltage
Output voltage and current
87
The block diagram of the linearized exciter
model with the conventional AVR is shown in
figure 8 [ 2, 4, 22, 23]: Where the parameters of
this linearized model are defined in Table 1
[1,2,24].
Fig. 8. Detailed Block Diagram of the Linearized Excitation System.
Table 1,
Description of the Excitation System Parameters.
Parameter Description Value Unit
K
D
Damping factor
= torque (pu) / speed (pu)
2 pu
T
M
Mechanical starting time 8 sec
K
A
Conventional AVR gain 50 -
T
A
Conventional AVR time
constant
0.02 sec
K
E
Exciter gain 0.17 -
T
E
Exciter time constant 0.95 sec
K
1
Synchronous Machine factor 1.0753
K
2
= 1.2581
K
3
= 0.3071
K
4
= 1.7124
K
5
= -0.0476
K
6
= 0.4972
T
3
Time constant of the field circuit 1.8 Sec
o
Frequency of the system 50 Hz
A
A
sT
K
+ 1

E E
sT K +
1
3
3
1 sT
K
+
K
2

M D
sT K +
1
K
1
V
ref
K
5
K
6
K
4
V
R
q
T
e
T
m
E
t
E
fd
-
-
-
+
s
o
Conventional
AVR
Exciter
Synchronous machine
88
The detailed block diagram of the excitation
control system in figure 8 is redrawn in a compact
form in figure 9 fitted with the proposed Digital
AVR(DAVR).
Fig. 9. Block Diagram of the linearized Excitation Control System with the Proposed Digital AVR (DAVR).
This configuration with Nbar scaling factor
(eq(15)) placed in the feedforward path (see figure
5) will be used to evaluate the proposed DAVR
for the excitation control system.
The continuous-time state-space notation
of the linearized open-loop system (Exciter +
Synchronous machine) can be written as in eq(1),
Where u is the control signal and y is the output.
Assuming that 0 = A
m
T , then:
R
V u A =
T
r
fd
fd
E E x
(
A A A A =
\

| |
T
r t
E y A A =

(
(
(
(
(
(
(
(
=
0 0 0
0
0
1
0 0 0
1 2
3
4 3
3 3
3
o
M M
D
M
E
E
T
K
T
K
T
K
T
K K
T T
K
T
K
A
(
(
(
(
(
(
=
0
0
0
1
E
T
B
(
=
0 1 0 0
0 0
5 6
K K
C
(
=
0
0
D
(17)
The above system (eq. (16)) is SIMO (Single
Input Multiple Output) and it represents the state-
space matrices of the exciter and synchronous
generator of figure 7.
4. Results and Simulations
The performance of the excitation system with
the conventional AVR is evaluated at different
values of the gain (K
A
). The MATLAB Control
Toolbox is utilized to obtain the results and the
simulations. Figures 10 -11 show the terminal
voltage (E
t
) and the angular speed deviation
(
r
) due to a step increase in the reference
voltage (V
ref
) by 0.3 pu.
Substituting the typical data of table 1 in
eq.(17) and calculating eq.(4) & eq.(5) with
sampling time sec 100 / 1 =
s
T , the discrete-time
state-space matrices of the open-loop plant (
d d
B A , ) are obtained:

(
(
(
(
=
0.9972 3.1111 0.0033 0
0.0018 - 0.9807 0.0021 - 0
0.0094 - 0.0148 - 0.9821 0.0055
0 0 0 0.9982
d
A
(
(
(
(
=
0
0
0
0105 . 0
d
B

(
= =
0 1 0 0
0.0476 - 0 0.4972 0
C C
d

(
= =
0
0
D D
d
With the matrix C given above, Q can be
determined by eq (11):
Exciter
-
Synchronous
machine
E
fd
E
t
DAVR
V
R
V
e
V
ref
Proposed AVR
89
(a)
(b)

(
(
(
(
=
0.0023 0.0237 - 0 0
0 0 0 0
0.0237 - 0 0.2472 0
0 0 0 0
Q
To get a best performance for the terminal
voltage (E
t
) and of the speed deviation (
r
),
the elements in main diagonal positions of Q will
be used to weight the outputs. Different values of
Q & R are selected by trial and error as mentioned
in table 2 below.
Fig. 10. Output Response of E
t
due to a 0.3 pu Step Increase in V
ref
, (a) K
A
= 10, (b) K
A
= 50.
Fig. 11. Output Response of
r
due to a 0.3 pu Step Increase in V
ref
, (a) K
A
= 30, (b) K
A
= 50.
(a)
(b)
90
Table 2:
State and control weishting matrices used in the
design .
With the first raw of Q & R in table 2 , the
digital LQR is calculated in (eq. (14)) and found
to be:

| | 0.6308 84.9536 - 17.4578 4.6571 = K
With the closed- loop poles calculated as the
eignvalues of ( K B A
d d
- ): 0.9713+j 0.0214,
0.9713-j0.0214, 0.9831+j0.0737, 0.9831-j0.0737,
all of the closed-loop eignvalues are inside the
unit circle, indicating that the system is
asymptotically stable. Time domain performance
of the DAVR is depicted in Figure 12, where the
terminal voltage (E
t
) and the angular speed
deviation (
r
) are plotted versus time. The
design of the considered values of Q and R of
table 2 (first raw) and an 0.3 pu step change in
the reference voltage ( V
ref
) are applied. It can
be seen from the output response of E
t
that the
settling time is 1 sec with 1.6% overshoot, while
the angular speed deviation
r
goes to zero in
just 2.35 sec.
The precision of the discrete system mainly
depends on the sampling time
s
T ; figure 13
emphasizes that as the value of T
s
gets higher, the
output response of the discrete systems gets closer
to the analog system's response with V
ref
= 0.5
pu. Last but not least, the response of the
DAVR's with various Q
s
and R
s
is shown in figure
14 with V
ref
= 0.6 pu.
Finally, it can be noticed that making values of
the Q's diagonal elements small leads to slow
response but with less overshoot, while large
values introduce large overshoot with swift
response. This can be figured out from figure 15.
Also, figure 15 describes the relationship between
R and the control signal u (
R
V A ), where large
value of R decreases
R
V A and hence the system
avoids actuator saturation. On contrast, the control
signal
R
V A goes higher in magnitude and may
exceed the allowable limit with small values of R
which turns the system into saturation.

Fig.12. Output Response of the Excitation Control System with DAVR due to a V
ref
= 0.3 pu , (a) E
t
, (b)
r
.
Q's R's
1
(
(
(
(
100 0.0237 - 0 0
0 100 0 0
0.0237 - 0 400 0
0 0 0 5 | | 1
2
(
(
(
(
1 0.0237 - 0 0
0 10 0 0
0.0237 - 0 5 0
0 0 0 5 | | 1 . 0
3
(
(
(
(
0.1 0.0237 - 0 0
0 1 0 0
0.0237 - 0 1 0
0 0 0 1 | | 3
(a)
(b)
91
(c)
(d)
(a)
(b)
(b)
Q1,R1
Q2,R2
Q3,R3
Q1,R1
Q2,R2
Q3,R3
(a)
Fig. 13. Output Response of the Excitation Control System ( E
t
) due to a V
ref
= 0.5 pu , (a) T
s
= 8 sec , (b) T
s
=16
sec , (c) T
s
= 45 sec, (d) T
s
= 85 sec.
Fig.14. Output Response of the Excitation Control System ( E
t
) due to a V
ref
0.6pu with Different DAVRs, (a)
E
t
, (b)
r
.
92
Fig. 15.Output Response of the Control Signal (V
R
) for Three Different Control Weighting Matrices (R).
5. Conclusion
In this paper a new optimal digital AVR for the
excitation control system is proposed based on the
linear quadratic techniques in order to control the
terminal voltage E
t
of the synchronous
generator. The performance of the proposed AVR
is evaluated on Single Machine Infinite Bus bar
(SMIB) system and compared with the
conventional AVR. A set of Q's and R's is
chosen and different DAVRs are obtained and
tested to restrict the output response of the system
within the desired range. Simulations show that in
addition to controlling E
t
, the proposed DAVR
damps well the angular speed deviations
r.
in
short time.
6. References
[1] Anderson P. M. and Fouad A. A., Power
System Control and Stability, John Wiley
and Sons Inc., 2003.
[2] Kundur P., Power System Stability and
Control, McGraw-Hill Inc., 1994.
[3] IEEE Committee Report, Computer
Representation of Excitation Systems,
IEEE Transactions on Power Apparatus and
Systems, Vol. Pas-87, No. 6, June 1968.
[4] Rangnekar S., Development of state-Space
and Study of Performance Characteristics of
Digital Based Excitation Control System
ST4B with Single Machine Connected to
Infinite Bus, IEEE International Conference
on Control Applications, Kohala Coast-
Island of Hawai'I, Hawai'I, USA, August 22-
27, 1999.
[5] Senjyu T., Yamashita T., Uezato K, and
Fujita H., LMI Approach Based Excitation
H Controller Design for Damping of Power
System Oscillations and Control of terminal
Voltage, IEEE International Conference on
Control Applications, 1999.
[6] Venayagamoorthy G. K. and Harley R. G.,
A Continually Online Trained
Neurocontroller for Excitation and Turbine
Control of a Turbogenerator, IEEE
Transactions on Energy Conversion, Vol. 16,
No. 3, September 2001.
[7] Venayagamoorthy G. K. and Harley R. G.,
Two Separate Continually Online-Trained
Neurocontrollers for Excitation and Turbine
Control of a Turbogenerator, IEEE
Transactions on Industry Applications, Vol.
38, No. 3, May/June 2002.
[8] Mitra, P., Chowdhury S.P., Pal S.K., and
Crossley P.A., Intelligent AVR and PSS
with Adaptive hybrid learning algorithm,
IEEE Power and Energy Society General
Meeting - Conversion and Delivery of
Electrical Energy in the 21st Century,
pages:1-7, 2008.
[9] Fusco G. and Russo M., A Adaptive
Voltage Regulator Design for Synchronous
Generator, IEEE Transactions on Energy
Conversion, Vol. 23 , Issue 3, pages: 946-
956, 2008.
[10] Mohammadi, S. M. A., Gharaveisi A. A.,
Mashinchi M., and Rafiei S. M. R., New
Evolutionary Methods For Optimal Design
Of PID Controllers for AVR System ,
93
IEEE Conference on Power Tech, pages 1-
8, Bucharest, Jun 28-Jul 2, 2009.
[11] Yoshimura S., Iki H., Uriu Y., Anai H.,
and Hyodo N., Generator excitation
control using a parameter space design
method ,43rd international universities
power engineering conference, pages 1-
4,Padova, Italy,1-4 September, 2008.
[12] Dysko A., Leithead W. E., O'Reilly J.,
Enhanced Power System Stability by
Coordinated PSS Design IEEE
Transactions on Power Systems, pages
413-422, 2010.
[13] Ogata K., Modern Control Engineering,
Prentice-Hall International, Inc, 1973.
[14] Control Tutorial for MATLAB and
Simulink, a website:
http://www.engin.umich.edu/class/ctms/ind
ex.htm
[15] Franklin G. F., Powell J. D., and Workman
M. L., Digital Control of Dynamic
Systems, Adison-Wesley, 3rd Edition,
1998.
[16] Fadali M. S., Digital Control Engineering:
Analysis and Design, Elsevier Inc, 2009.
[17] Ogata K., Modern Control Engineering,
4th edition, Prentice-Hall International, Inc,
2002.
[18] Moudgalya K. M., Digital Control, John
Wiley & Sons, Ltd, 2007.
[19] Control System Toolbox, MATLAB, The
MathWorks, Inc, 2009.
[20] Dutton K., Thompson S., and Barraclough
B., The Art of Control Engineering ,
Adison-Wesley,1997.
[21] Law B. E., Simulation of the transient
response of synchronous machines, B.Sc.
thesis, the university of Queensland, School
of Information Technology and Electrical
Engineering, October, 2001.
[22] Chen S. and Malik O. P., Power System
Stabilizer Design using Synthesis, IEEE
Transactions on Energy Conversion, Vol.
10, No. 1, March 1995.
[23] Franklin G. F., Powell J. D., and Workman
M. L., A Robust H Synthesis Controller
for Power System Stabilization,, ASCE
Journal, Volume (5), Issue (II), June, 2005.
[24] IEEE Power Engineering Society, IEEE
Recommended Practice for Excitation
System Models for Power System Stability
Studies, IEEE Inc., 21 April 2006.

7 1 94 - 82 ) 2011 (
94

/ /
: [email protected]

. R Q .

. )
( . ) (
.


A Digital-Based Optimal AVR Design of Synchronous Generator Exciter Using LQR Technique

Uploaded by

Copyright:

Available Formats

A Digital-Based Optimal AVR Design of Synchronous Generator Exciter Using LQR Technique

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Digital-Based Optimal AVR Design of Synchronous Generator Exciter Using LQR Technique

Uploaded by

Copyright:

Available Formats

Al-Khwarizmi

You might also like