Engineering Science and Technology, an International Journal 21 (2018) 1104–1111

Contents lists available at ScienceDirect

Engineering Science and Technology,

an International Journal
journal homepage: www.elsevier.com/locate/jestch

Full Length Article

Performance enhancement of automatic voltage regulator by modified

cost function and symbiotic organisms search algorithm
Emre Çelik a,⇑, Rafet Durgut b
a
Electrical and Electronics Engineering, Engineering Faculty, Düzce University, Düzce, Turkey
b
Computer Engineering, Engineering Faculty, Karabük University, Karabük, Turkey

a r t i c l e i n f o a b s t r a c t

Article history: This article attempts to solve the problem of efficient design of proportional + integral + derivative (PID)
Received 26 May 2018 controller applied to popular automatic voltage regulator (AVR) system by employing recently intro-
Revised 19 July 2018 duced symbiotic organisms search (SOS) algorithm, for the first time. PID controller design needs proper
Accepted 9 August 2018
determination of three control parameters. Such a design problem can be taken as an optimization task
Available online 16 August 2018
and SOS is invoked to find out better controller parameters through a new cost function defined in the
paper, which allows to evaluate the control behavior in both time-domain and frequency-domain. For
Keywords:
the performance analysis, distinct analysis techniques are deployed such as transient response analysis,
Automatic voltage regulator
PID controller
root locus analysis and bode analysis. Besides, robustness analysis of the closed-loop control system
Multi-objective optimization tuned by SOS is performed with regard to parameter uncertainties and external disturbance. The efficacy
Symbiotic organisms search algorithm of the presented technique is widely illustrated by comparing the obtained results with those reported in
Performance analysis some prestigious journals and it is shown that our proposal leads to a more satisfactory control perfor-
Cost function mance from the perspective of both time-domain and frequency-domain specifications while with a good
robustness to parameter uncertainties and unknown changes in the system output.
Ó 2018 Karabuk University. Publishing services by Elsevier B.V. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction control, where the exciter voltage is regulated to match the voltage
drop or rise according to the new conditions [3].
In an electric power network, ensuring a constant voltage level In the hope of implementing and enhancing dynamic response
under various circumstances is one of the significant control issues of an AVR system, several control techniques have been studied
of power systems having a close relation to power quality, grid in the literature based on optimal control, robust control, fuzzy
security and grid reliability. When facing a deviation in grid volt- logic, conventional and fractional order proportional + integral +
age level, it leads to remarkable changes in the system dynamics derivative (PID) techniques and adaptive control, which have indi-
and accordingly there may be a deterioration in the performance vidual advantages and disadvantages. Among the reported con-
of the devices connected with this power grid and drop in their life trollers, the classical PID is no doubt the one that is the most
expectancy, because all equipment can operate efficiently only for preferred owing to its robust performance regardless of variations
a particular voltage level termed as nameplate or rated voltage in system parameters and structural simplicity which requires tun-
[1,2]. Moreover, controlling the bus voltage in a local sense has ing of only three control parameters, such as proportional gain,
another aspect of regulating reactive power flow, thus rendering integral gain and derivative gain [4]. However, proper determina-
it possible to reduce real line losses because of the reactive current tion of PID gains is fairly difficult and there is no universal method-
components in electric power network. In order to fulfill the afore- ology that assists the operator in designing this controller. When
said objectives, automatic voltage regulator (AVR) system is the literature is evaluated, it can be seen that a vast number of arti-
installed in electrical power systems. An AVR is equipment aimed ficial intelligence algorithms have been paid much attention by
to sustain the output voltage of a synchronous generator (SG) at a many researchers particularly since 2000, so as to acquire an
desirable voltage level by keeping its excitation voltage under almost optimal solution in their PID design by exploiting the
unique search ability of governed optimization algorithm and/or
their cost function definition. In this context, in 2012, artificial
⇑ Corresponding author. bee colony (ABC) algorithm is suggested to enhance the perfor-
E-mail address: [email protected] (E. Çelik). mance of PID-controlled AVR system, where a comparison with
Peer review under responsibility of Karabuk University.

https://doi.org/10.1016/j.jestch.2018.08.006
2215-0986/Ó 2018 Karabuk University. Publishing services by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 E. Çelik, R. Durgut / Engineering Science and Technology, an International Journal 21 (2018) 1104–1111 1105

particle swarm optimization (PSO) and differential evolution (DE) model time constants within the range of +50% to
50% in steps
algorithm are also presented [1]. From the findings, ABC is found of 25% and also in the face of external disturbances in the system
to exhibit better performance than the others. Subsequently, many output.
optimizing liaisons (MOL) algorithm, which is the simplified revi-
sion of the original PSO, is applied to the same optimization prob-
lem of searching for better PID parameters [2]. The results are 2. PID controller-based AVR design
compared to those in [1] and it is shown that MOL-based PID
controller can enhance the system performance with regard to In spite of many efforts in control engineering field, PID con-
both time-domain and frequency-domain measures. In 2016, troller or its cousins have been still widely used in various types
biogeography-based optimization (BBO) algorithm is introduced of control systems [10]. The reason of this wide usage comes from
into searching for optimal PID parameters for the concerned con- its easily understandable nature, ease of design and robust perfor-
trol system [5]. Comparative results with ABC-based obtained mance irrespective to model uncertainties with proper tuning of
results in [1] demonstrate that BBO algorithm outperforms the controller parameters [11]. In s-domain, the transfer function of a
ABC approach, thereby it yields an improvement in the system PID controller is expressed by
dynamic response. In [6], PSO and global neighborhood algorithm
U ðsÞ Ki
(GNA) are adopted to optimize the output response of a PID- GPID ðsÞ ¼ P þ I þ D ¼ ¼ Kp þ þ Kds ð1Þ
EðsÞ s
controlled AVR system. From the results of transient response
analysis, GNA is found to perform better than PSO with regard to
where EðsÞ is the error variable between the desired and real pro-
settling time and rise time. However, peak overshoot of the
cess output which produces the control signal U ðsÞ by computing
response with GNA is greater than that of PSO. An application of
the sum of proportional term P, integral term I and derivative term
chaotic PSO (CPSO) is made in [7] to optimize the AVR system per-
D. The three design parameters of this controller, i.e. proportional
formance. A comparison is also presented with the results obtained
gain K p , integral gain K i and derivative gain K d , must be tuned
by the standard PSO in [6]. It is shown that CPSO-based AVR
jointly by the operator depending upon the plant’s dynamics. The
system performance is improved considering peak overshoot and
resulting response against a unit step input should engage with
settling time. However, it can be said that the validation of these
the given reference with minimal settling time and no sustained
two studies [6,7] is not properly justified because no published
oscillation.
work is used for comparison.
In an electric power grid, there are more than one generator
Symbiotic organisms search (SOS) algorithm is a relatively
connected to similar busbar and each has its own AVR. As previ-
straightforward and effective metaheuristic proposed by Cheng
ously mentioned, the design objective of an AVR is to sustain the
and Prayogo in 2014 [8]. In the algorithm, simulation of symbiotic
output voltage of a SG at a certain level. As pictured in Fig. 1, an
interaction strategies observed amongst organisms in order to
AVR includes mainly four essential components, such as amplifier,
keep alive in the ecosystem is realized. A significant advantage of
exciter, generator and sensor. In this system, as the aim is to con-
the algorithm is that it requires only two common tuning parame-
trol the voltage of power utility that the generator is connected to
ters such as population size and maximum iteration number. Pre-
via power transformer, the voltage level is continuously measured
liminary tests of applying SOS to some mathematical benchmark
as feedback signal using a voltage sensor. After being rectified and
problems and engineering design problems affirm the excellence
filtered out, this signal is compared to the voltage setpoint in the
of the SOS compared with other remarked optimization algo-
comparator in order to obtain voltage error signal. The error signal
rithms. In addition, superior performance of the SOS for optimizing
is amplified, then is fed to exciter to adjust the generator field
PI parameters in an off-line sense for a DC servo motor drive is
winding voltage/current so that any deviation in generator termi-
demonstrated based on simulated and experimental results in [9]
nal voltage resulting from new operating conditions could be com-
as compared to PSO, genetic algorithm and classical Ziegler-
pensated in a quick and stable behavior.
Nichols tuning rule. To the authors’ knowledge, it has not been
In order to investigate the AVR dynamic performance mathe-
yet addressed in the open literature whether the application of
matically, the following transfer function modelling is assumed,
SOS leads to more optimal PID controller gains or not in presence
where the major time constants are used and saturation or other
of AVR control application.
nonlinearities are avoided in the way similar to the literature stud-
In the light of the consequences of the above paragraphs, the
ies [1–7,10,12–15].
authors of this article are encouraged to present a unique design
methodology for the studied AVR system that improves the
trade-off between the dynamic response and the stability margin comparator
of the system, which is, as figured out by the earlier works, in an Turbine + Voltage
vdc Exciter Amplifier
error
Σ setpoint
insufficient level in the literature. To fill this research gap, the +
signal
design problem is contemplated as an optimization task and a
Field
new composite cost function in the time-domain and frequency- F1 winding F2
domain is suggested. Then, SOS is invoked to optimize the PID con- Shaft Rectifier
troller gains so that the controlled system may yield the aspired Electric Generator & Filter
response and degree of stability as depicted by the suggested cost (Alternator)
function. Using transient response analysis, root locus analysis and
bode analysis, the performance of presented AVR system is widely
established in comparison with those based on ABC [1], MOL [2] Power Voltage
and BBO [5]. The extensive results reported in this article show transformer sensor
that the output voltage profile settles to the unit step reference
with the least peak overshoot without compromising on settling
time much. This outcome has improved the stability margin of
the AVR system compared to other reported approaches. In order Power Utility
to complement the contribution of this study, robustness of the
presented controller is also validated under the variations of the Fig. 1. Schematic diagram of an AVR system.
1106 E. Çelik, R. Durgut / Engineering Science and Technology, an International Journal 21 (2018) 1104–1111

A. Amplifier model: The amplifier model is given by a gain K A and 1.6

a time constant sA , as below. X: 0.753

1.4 Y: 1.507

KA
GAmplifier ðsÞ ¼ ð2Þ
1 þ sA s 1.2

change of V (s) (V)

where K A can vary in the range of 10–40 while sA ranges between 1

0.02 s and 0.1 s.

t
X: 17.31 X: 20
0.8 Y: 0.9091 Y: 0.9091

B. Exciter model: Like amplifier, transfer function model of an

exciter may be represented by a gain K E and a time constant sE 0.6
and is given by
0.4
KE
GExciter ðsÞ ¼ ð3Þ
1 þ sE s 0.2

Standard values of K E are in the range of 1–10 and sE in the 0

0 5 10 15 20
range of 0.4–1.0 s.
time (s)
C. Generator model: The generator is modelled by a gain K G and a
time constant sG , as presented in Eq. (4). Fig. 3. Original terminal voltage step response change of an AVR system.

KG
GGenerator ðsÞ ¼ ð4Þ
1 þ sG s
Herein, K G and sG are the constants dependent on generator
loading conditions. K G ranges from 0.7 to 1.0 and sG is between
1.0 s and 2.0 s.
D. Sensor model: The sensor circuit which is responsible for
measuring, rectifying and smoothing the system voltage is often
modelled by a gain K S and a time constant sS , as given in Eq. (5).
KS
GSensor ðsÞ ¼ ð5Þ
1 þ sS s Fig. 4. Transfer function block diagram of the PID controller-based AVR design
including a disturbance.
where sS normally takes small values ranging over 0.001–0.06 s and
K S is in the neighborhood of 1.0.
In this article, to lead a fair comparison with [1,2,5], the same controller-based AVR design which also adds a disturbance signal
parameter values have been used as K A ¼ 10, sA ¼ 0:1, K E ¼ 1:0, to DV t ðsÞ is shown in Fig. 4.
sE ¼ 0:4, K G ¼ 1:0, sG ¼ 1:0, K S ¼ 1:0 and sS ¼ 0:01. Adopting the Eventually, assuming the disturbance value in Fig. 4 is zero, the
above parameter values of the model, the entire AVR transfer final transfer function model of the AVR system employing a PID
function block diagram is given in Fig. 2. GAVRP ID ðsÞ can be represented by
From Fig. 2, the system transfer function GAVR ðsÞ could be
extracted as in Eq. (6). GPID ðsÞ  GAmplifier ðsÞ  GExciter ðsÞ  GGenerator ðsÞ
GAVRP ID ðsÞ ¼ ð7Þ
1 þ GPID ðsÞ  GAmplifier ðsÞ  GExciter ðsÞ  GGenerator ðsÞ
DV t ðsÞ 0:1s þ 10
GAVR ðsÞ ¼ ¼
DV ref ðsÞ 0:0004s4 þ 0:0454s3 þ 0:555s2 þ 1:51s þ 11 While we recognize that the system model is not very complex,
ð6Þ it has been used popularly in its current state for many years to
introduce incremental improvements. As a result of our literature
Notice that the input DV ref ðsÞ and output DV t ðsÞof this system review, it is noticed that there may be still a research gap in
are not current values of the reference input and terminal voltage improving the system performance further. In this regard, the con-
quantities, but stand for the incremental changes of the corre- tribution of this article is not to propose a more realistic AVR model
sponding variables. Using Eq. (6), the original terminal voltage step nor to verify a novel scheme for controlling AVR, but to effectively
response change of the above AVR system is depicted in Fig. 3, from bridge that research gap in the hoping of enhancing the time-
which the system is observed to be severely oscillating in the domain and frequency-domain performance of the existing AVR
beginning and has remarkable error at steady-state. In power sys- system using similar PID controller except that its design parame-
tems, such a response is completely unacceptable and cannot be ters are tuned with the guidance of a new cost function which is to
allowed to emerge when considering the operating voltage in the be minimized by the employment of SOS algorithm.
order of kilo-volts.
To improve the transient response of the AVR system and elim-
inate the steady-state error, a controller such as a PID is required to 3. Design of PID employing SOS
be installed in the concerned system. Block diagram of PID
In this part, an efficient PID controller design employing the SOS
algorithm, which can be shortly referred to as SOS-PID controller, is
realized to enhance the voltage response profile of the AVR system
while maintaining satisfactory stability margin. The incorporation
of SOS in such a problem is primarily owing to the desire of attain-
ing three controller coefficients K p , K i and K d , so that the controlled
system can have the desired performance. SOS is a relatively new
algorithm proved to be powerful and robust over different kinds
of optimization problems [8]. It operates on the basis of three com-
mon symbiotic strategies such as mutualism, commensalism, and
Fig. 2. Transfer function block diagram of the AVR system. parasitism developed by organisms. The property of the interaction
 E. Çelik, R. Durgut / Engineering Science and Technology, an International Journal 21 (2018) 1104–1111 1107

characterizes the basis of each phase. Mutualism is the symbiotic

Initial ecosystem
interaction in which each organism benefits from the other’s activ- generated randomly
ity. Commensalism evolves when one organism receives benefits,
while the other organism is neutral, and parasitism evolves when Compute fitness set iter = iter + 1 Identify best
an organism gains benefits from a certain interaction at the cost by Eq.8 and i = 1 organism
of degrading the other [16]. Through all phases, each organism
interacts randomly with other organisms in the ecosystem. After i=i+1
completion of these three symbiotic strategies, SOS tries another yes Modification
generation and is iterated recursively until pre-defined termina- i ≤ eco_size ?
Performance Mutualism
evaluation Commensalism
tion criteria are satisfied. The following outline summarizes the no Parasitism
afore-said explanations. no
Is termination
Initialization criterion met? PID controlled AVR model
REPEAT
yes
i = 1; AVR
Optimized PID Σ PID controller
while i is different from eco_size plant
parameters

Mutualism phase

Commensalism phase
Fig. 5. SOS program implementation for optimizing PID controller gains in AVR

Parasitism phase control application.

i = i + 1;
end while
organism has K p , K i and K d gains ranging from 0.01 to2.0. After
UNTIL (termination criteria are met)
the best organism is identified, organisms are used in the PID con-
For further insight into the SOS algorithm procedures, readers
trol law in order to simulate the system behavior by means of the
are referred to the original study given by [8].
PID controlled AVR model. As predicted, each organism exhibits
In order to implement SOS algorithm for optimizing the PID
different terminal voltage curve with its own time-domain
controller gains, three design parameters are initially defined to

 response and frequency-domain response. Then, using Eq. (8), a fit-
form an individual organism K by K ¼ K p ; K i ; K d , where each
ness value is computed for each organism. Later on, the organisms
member is represented by a real number. Thus, there are three
are modified using the particular SOS phases for the next
members in an individual to be optimized and each individual
iterations.
may be treated as a PID controller with different gains. To assess
This process of identifying best organism, performance evalua-
the performance of distinct PID controllers in the ecosystem, a suit-
tion and employing SOS strategies is iterated until i is equal to the
able cost function complying with the requirements and necessi-
number of organisms in the ecosystem. Otherwise, unless termina-
ties in a control system design must be defined properly. In
tion criterion is met, iteration number is increased and i is set back
literature, there exist various performance measures for design of
to 1 again, and the whole process is repeated. The PID parameters
controllers such as integral of absolute error (IAE), integral of
obtained at the end of the program are used in the subsequent
squared error (ISE) and integral of time weighted squared error
simulations.
(ITSE) [17]. An important deficiency of IAE and ISE is that they lead
to sluggish response due to the accumulated errors regardless of
the time. On the other hand, ITSE can solve this problem and 4. Numerical results
increase the dynamic response, but this is not desirable as far as
stability margin is concerned as in [3]. Instead, integral of time In this section, simulation results obtained after applying the
weighted absolute error (ITAE) has been demonstrated in several presented technique to the AVR system are provided, and a fair
studies to exhibit better system performance as compared to its comparison from perspective of transient response analysis, root
integral-based counterparts [18]. In order to make this study com- locus analysis and bode analysis is also presented in comparison
pete with [1,2,5], two more parameters obtained from the with ABC [1], MOL [2] and BBO [5], which have been published
frequency-domain response of the system are considered and com- in esteemed journals. In SOS algorithm, only two parameters are
bined with the ITAE criterion by weighting each term as shown in set as ecosystem size = 30 and maximum iteration number = 30. Sim-
Eq. (8). ulations were implemented in the Matlab 8.5.0 (R2015a) software
Z installed on a computer with an Intel core (TM) i5 3.3 GHz proces-
t sim
1:0 sor and 8 GB memory.
J ¼ x1 t jDV e ðt Þjdt þ x2 a þ x3 ð8Þ
b One of the major observations of the present study for analyzing
0
the voltage response curve is given in Fig. 6, where the disturbance
The first term on the right side of Eq. (8) relates to ITAE which value in Fig. 4 is considered zero. In this figure, the change of out-
accumulates the product of the current time t and absolute error in put voltage response of proposed PID-controlled AVR system to a
terminal voltage jDV e ðt Þj up to the sufficiently chosen simulation step command is portrayed comparatively with other indicated
time tsim . a is the number of complex poles computed from the approaches. From Fig. 6, it is apparently viewed that the systems
characteristic equation of the system while b is the sum of the are poorly damped with ABC- and BBO-tuned PID controllers. It
damping ratios of the complex poles. Importance of each term in is also noticed that as compared to MOL-based response, the
J is set by a weight factor xj . Recommended values of the weight designed controller offers less peak overshoot (black trace) while
factors are x1 ¼ 0:71, x2 ¼ 0:2 and x3 ¼ 0:09. It is worth high- maintaining nearly the same settling time, which is a prior indica-
lighting that better performance is got when the value of J is min- tor for improvement of stability degree of the concerned control
imum which requires that the values of ITAE and a are minimized application. Also notice that the responses with ABC and BBO can-
and that of b is maximized. For increased stability margin, a less not be hold at 1.0 pu for the considered simulation time. This is
number of complex poles and their associated damping ratios clo- attributed to the use of ITSE cost function in those studies, which
ser to one are more preferred. leads to unrealistic evaluations owing to squaring error. As a result,
In the presented technique visualized in Fig. 5, initial ecosystem a better setpoint tracking performance is achieved with the pre-
including a number of organisms is generated randomly; each sented technique than other techniques.
1108 E. Çelik, R. Durgut / Engineering Science and Technology, an International Journal 21 (2018) 1104–1111

tions of Table 1. In the reported results, bold text indicates com-

paratively the best result.
It is noticeable From Table 1 that SOS-based PID control offers
the best value of peak overshoot and comparable settling time with
only 2.9% less than MOL-based result. On the other hand, steady-
state error values of SOS- and MOL-based results are the same,
and both are significantly better that those of ABC and BBO. The
best result concerning rise time belongs to BBO algorithm.
For the stability concern of the studied AVR system optimized
by the proposed technique, root locus analysis is performed and
the respective root locus curve is depicted in Fig. 7. As shown, all
the closed-loop poles are located at the left side of the s-plane,
meaning that the proposed control application is stable.
The closed-loop poles and their respective damping ratios in
Fig. 7 are also computed and gathered in Table 2 in comparison
with the other indicated studies. It is clear that the conjugate poles
of the presented AVR system are farther away from the imaginary
axis, which makes the system be controlled with the biggest damp-

Fig. 6. Comparative terminal voltage changing profiles.

Table 2
Closed-loop poles and their respective damping ratios of the AVR system optimized
The transient response and steady-state performances regard-
by SOS, ABC, MOL and BBO algorithm.
ing the time responses are measured from Fig. 6 and reported in
Table 1. Transient response analysis covers the time domain per- Algorithm Closed-loop pole Damping ratio

formance characteristics such as peak overshoot (M P ), settling time Presented
100.48 1

(T S , 5% band) and rise time (T R ) whereas steady-state performance
1.98 1

1.10 1
is with the error value at steady-state (ESS ). In addition, optimized

4.97 + 4.69i 0.727
PID controller parameters are also given in their respective sec-
4.97
4.69i 0.727
ABC [1]
100.98 1

4.74 1

0.25 1
Table 1

3.75 + 8.40i 0.40
Comparative controller parameters and corresponding system performance

3.75
8.40i 0.40
specifications.
MOL [2]
100 1
Controller parameters/ Presented ABC [1] MOL [2] BBO [5]
2.11 1
performance/techniques
1.06 1
Kp 0.5693 1.6524 0.5857 1.2464
4.92 + 4.72i 0.721
Ki 0.4097 0.4083 0.4189 0.5893
4.92
4.72i 0.721
Kd 0.1750 0.3654 0.1772 0.4596 BBO [5]
100.0 1
MP 1.013 1.250 1.020 1.160
2.1 1
TS 0.485 0.920 0.471 0.766
0585 1
TR 0.353 0.156 0.343 0.149
4.8 + 10.2i 0.427
ESS 0.002 0.026 0.002 0.014
4.8–10.2i 0.427

Fig. 7. Root locus curve of the studied AVR system. Fig. 8. Bode diagram of the proposed AVR system.
 E. Çelik, R. Durgut / Engineering Science and Technology, an International Journal 21 (2018) 1104–1111 1109

ing ratio value of 0.727, which is 45% more than ABC, 0.83% more is achieved by deploying SOS and MOL, and they are the pioneers
than MOL and 41% more than BBO. over the remaining techniques.
So as to investigate the stability of the proposed AVR through Now that the presented AVR system arguably performs better
another point of view, frequency response or bode analysis of the than the existing studies, the paper is extended by only focusing
control system is conducted, and the resulting bode diagram is on the robustness of SOS optimized PID controller in presence of
depicted in Fig. 8. The peak gain, phase margin, delay margin and model uncertainty and external disturbances. Such a robustness
bandwidth parameter corresponding to this bode plot are tabu- analysis, which is previously missing in the conference version of
lated in Table 3. this article, is indispensable as the final stage to validate any novel
From Table 3, it is seen that the minimum peak gain, maximum control scheme. To bridge this gap, first, time constants of the sys-
phase margin and maximum delay margin, which are essential fac- tem elements such as amplifier, exciter, generator and sensor are
tors required for enhanced stability, are provided by both our pro- changed separately in the range ±50% of the nominal value in steps
posal and that based on MOL algorithm. With regard to bandwidth, of 25% while the controller parameters in Table 1 remain the same.
its maximum value is offered by using BBO algorithm. As a conse- The resulting responses to sudden changes in sA , sE , sG and sS are
quence, as far as peak gain, phase margin and delay margin mea- shown by four subplots in Fig. 9 along with the nominal response
sures of the bode analysis are concerned, the same performance and the reference input signal.
Numerical results of the robustness analysis obtained from
transient response analysis applied to the plots in Fig. 9 are also
computed and tabulated in Table 4 for each time constant
Table 3 parameter.
Peak gains, phase margins, delay margins and bandwidths of different AVR systems
Moreover, the range of total deviation between the maximum
using SOS, ABC, MOL and BBO algorithm.
and minimum value of a time-domain performance parameter,
Presented ABC [1] MOL [2] BBO [5] and the percentage of maximum deviation relative to the corre-
Peak gain (dB) 0.0 2.87 0.0 1.56 sponding nominal value are also reported in Table 5 in order to
Phase margin (deg.) 180 69.4 180 81.6 give a better picture of the robustness analysis. It is clear from
Delay margin (s) Inf. 0.111 Inf. 0.122
Table 5 that the deviations from the nominal values are generally
Bandwidth 6.15 12.88 6.34 14.28
small. As we can see at first glance, variation of the sensor time

Fig. 9. Step responses of the studied AVR system under the variation of time constants (a) sA (b) sE (c) sG (d) sS .
1110 E. Çelik, R. Durgut / Engineering Science and Technology, an International Journal 21 (2018) 1104–1111

Table 4
Numerical results of robustness analysis of the AVR system controlled by SOS
optimized PID controller.

Parameter Rate of Peak Settling Rise

variation (%) Overshoot (V) Time (s) Time (s)
sA +50 1.072 0.905 0.355
+25 1.043 0.481 0.351

25 1.001 0.519 0.367

50 1.001 0.607 0.408
sE +50 1.078 1.336 0.431
+25 1.047 0.534 0.393

25 0.998 0.438 0.307

50 0.994 0.958 0.258
sG +50 1.053 1.434 0.489
+25 1.030 0.578 0.423

25 1.000 0.384 0.278

50 0.999 1.017 0.202
sS +50 1.019 0.472 0.345
+25 1.016 0.479 0.348

25 1.010 0.492 0.357

50 1.008 0.499 0.362 Fig. 10. Step responses of different controllers under external disturbance.

5. Conclusion
Table 5
Range of total deviations and percentage of maximum deviations (%) under parameter
Tuning problem of control parameters of a PID controller work-
uncertainty. ing in an AVR control application is addressed and tried to be
solved in a better fashion by introducing a new cost function which
Time Parameter Range of total Percentage of maximum
constant deviations deviations (%)
is optimized employing SOS algorithm. After the cost function def-
inition is realized with regard to both time-domain and frequency-
sA Peak overshoot (V) 0.071 5.82
Settling time (s) 0.424 86.6
domain performance criteria, SOS is invoked subsequently as the
Rise time (s) 0.057 15.6 powerful optimization technique to tune controller gains in a sense
sE Peak overshoot (V) 0.084 6.4
that minimum cost function value could be achieved. In order to
Settling time (s) 0.898 175.5 appraise the effectiveness of presented approach, three popular
Rise time (s) 0.173 26.9 studies are chosen from the literature as a benchmark, then the
sG Peak overshoot (V) 0.054 3.9 results are compared under identical conditions from the perspec-
Settling time (s) 1.050 195.7 tive of diverse analysis techniques such as transient response anal-
Rise time (s) 0.287 42.8 ysis, root locus analysis and bode analysis. Simulation results show
sS Peak overshoot (V) 0.011 0.6 that the cooperation of the developed cost function and SOS algo-
Settling time (s) 0.027 2.89 rithm improves the trade-off between the dynamic response and
Rise time (s) 0.017 2.5
the stability margin of the system. In this context, the presented
approach is able to effectively improve the stability degree of the
considered AVR system by further reducing the peak overshoot
of the system time response compared to existing approaches.
constant sS has almost negligible impact over the system response Moreover, depending upon the results evolved from the robustness
compared to that of other time constants. Given an example, fol- analysis, it is found that the proposed AVR system is well capable
lowing a change in sA in the specified interval, peak overshoot to maintain the desired response when exposed to the applied
deviates from its nominal value in the range 0.071 V, settling time parameter uncertainties and external disturbance. Finally, accord-
in the range 0.424 s and rise time in the range 0.057 s, which lead ing to the various tests performed by the authors, it has been seen
to maximum deviations of 5.82%, 86.6% and 15.6% with respect to that if desired, stability margin of the system can be further
the nominal values, respectively. Considering all the time con- improved by increasing x3 in Eq. (8). However, this will degrade
stants, the average deviation of the peak overshoot, settling time the transient-time characteristics such as rise time and settling
and rise time are 4.2%, 115.2% and 22%, respectively. The fact that time. For interested researchers, it would be of interest to use
all the ranges of total deviations are approximately below 0.5 any other powerful optimization algorithms along with the cost
proves that the SOS-PID controller is robust and preserves the function defined in this paper in the hope of improving the AVR
desired transient response regardless of the variation in any of performance further.
the time constants in the considered change interval.
Finally, in order to verify disturbance rejection capability of our
contribution against other techniques, an external disturbance is
References
introduced between t = 3 and t = 5 s by setting the disturbance
value in Fig. 4 as 0.15. As is observed in Fig. 10, in all cases the ter- [1] H. Gozde, M.C. Taplamacioglu, Comparative performance analysis of artificial
minal voltage changing curve settles to the reference value after bee colony algorithm for automatic voltage regulator (AVR) system, J. Franklin
Inst. 348 (2011) 1927–1946.
each perturbation; however, responses to external disturbance
[2] S. Panda, B.K. Sahu, P.K. Mohanty, Design and performance analysis of PID
using ABC and BBO exhibit some oscillations and longer settling controller for an automatic voltage regulator system using simplified particle
time. On the other hand, SOS- and MOL-PID controllers gain a sim- swarm optimization algorithm for automatic voltage regulator (AVR) system, J.
ilar level of disturbance rejection where the disturbances owing to Franklin Inst. 349 (2012) 2609–2625.
[3] E. Çelik, Incorporation of stochastic fractal search algorithm into efficient
the unknown changes in the system output are rejected well, but design of PID controller for an automatic voltage regulator system, Neural
with a slightly faster sense in the case of our proposal. Comput. Appl. (2018), https://doi.org/10.1007/s00521-017-3335-7.
 E. Çelik, R. Durgut / Engineering Science and Technology, an International Journal 21 (2018) 1104–1111 1111

[4] S. Chatterjee, V. Mukherjee, PID controller for automatic voltage regulator [12] P.K. Mohanty, B.K. Sahu, S. Panda, Tuning and assessment of proportional–
using teaching–learning based optimization technique, Int. J. Electr. Power integral–derivative controller for an automatic voltage regulator system
Energy Syst. 77 (2016) 418–429. employing local unimodal sampling algorithm, Electr. Power Compo. Syst.
_ Akkaya, Performance analysis of biogeography-
[5] U. Güvenç, T. Yiğit, A.H. Isßık, I. 42 (2014) 959–969.
based optimization for automatic voltage regulator system, Turk. J. Electr. Eng. [13] H. Shayeghi, A. Younesi, Y. Hashemi, Optimal design of a robust discrete
Comput. Sci. 24 (2016) 1150–1162. parallel FP + FI + FD controller for the Automatic Voltage Regulator system, Int.
[6] H. Gözde, M.C. Taplamacioglu, M. Ari, Simulation study for global J. Electr. Power Energy Syst. 67 (2015) 66–75.
neighborhood algorithm based optimal automatic voltage regulator (AVR) [14] P.S. Bandaghiri, N. Moradi, S.S. Tehrani, Optimal tuning of PID controller
_
system Istanbul, in: 5th International Istanbul Smart Grids and Cities Congress parameters for speed control of dc motor based on world cup optimization
and Fair, 2017, pp. 46–50. algorithm, Tech. J. Eng. Appl. Sci. 6 (2016) 106–111.
[7] H. Gözde, M.C. Taplamacioglu, M. Ari, Automatic voltage regulator (AVR) [15] A. Chatterjee, V. Mukherjee, S.P. Ghoshal, Velocity relaxed and craziness-based
design with chaotic particle swarm optimization, in: International Conference swarm optimized intelligent PID and PSS controlled AVR system, Int. J. Electr.
on Electronics, Computers and Artificial Intelligence, 2014, pp. 23–26. Power Energy Syst. 31 (2009) 323–333.
Bucharest, Romania. [16] V.F. Yu, A.A.N.P. Redi, C.L. Yang, E. Ruskartina, B. Santosa, Symbiotic organisms
[8] M.Y. Cheng, D. Prayogo, Symbiotic organisms search: a new metaheuristic search and two solution representations for solving the capacitated vehicle
optimization algorithm, Comput. Struct. 139 (2014) 98–112. routing problem, Appl. Soft Comput. 52 (2017) 657–672.
[9] E. Çelik, N. Öztürk, First application of symbiotic organisms search algorithm to [17] M. Zamani, M.K. Ghartemani, N. Sadati, M. Parniani, Design of a fractional
off-line optimization of PI parameters for DSP-based DC motor drives, Neural order PID controller for an AVR using particle swarm, Control Eng. Pract. 17
Comput. Appl., doi: 10.1007/s00521-017-3256-5. (2009) 1380–1387.
[10] L.S. Coelho, Tuning of PID controller for an automatic regulator voltage system [18] S. Padhy, S. Panda, A hybrid stochastic fractal search and pattern search
using chaotic optimization approach, Chaos Solitons Fractals 39 (2009) 1504– technique based cascade PI-PD controller for automatic generation control of
1514. multi-source power systems in presence of plug in electric vehicles, CAAI
[11] D. Guha, P.K. Roy, S. Banerjee, Study of differential search algorithm based Trans. Intelligence Technol. 2 (2017) 12–25.
automatic generation control of an interconnected thermal-thermal system
with governor dead-band, Appl. Soft Comput. 52 (2017) 160–175.