Engineering Science and Technology, an International Journal 18 (2015) 194e206

H O S T E D BY Contents lists available at ScienceDirect

Engineering Science and Technology,

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

Full length article

A novel optimal PID plus second order derivative controller

for AVR system
Mouayad A. Sahib
Software Engineering Department, Salahaddin University, Kurdistan Regional, Erbil, Iraq

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

Article history: This paper proposes a novel controller for automatic voltage regulator (AVR) system. The controller is a
Received 6 August 2014 four term control type consisting of proportional, integral, derivative, and second order derivative terms
Received in revised form (PIDD2). The four parameters of the proposed controller are optimized using particle swarm optimization
7 November 2014
(PSO) algorithm. The performance of the proposed PIDD2 is compared with various PID controllers tuned
Accepted 19 November 2014
Available online 6 January 2015
by modern heuristic optimization algorithms. In addition, a comparison with the fractional order PID
(FOPID) controller tuned by Chaotic Ant Swarm (CAS) algorithm is also performed. Furthermore, a fre-
quency response, zero-pole map, and robustness analysis of the AVR system with PIDD2 is performed.
Keywords:
Optimal control
Practical implementation issues of the proposed controller are also addressed. Simulation results showed
PID controller a superior response performance of the PIDD2 controller in comparison to PID and FOPID controllers.
Automatic voltage regulator Moreover, the proposed PIDD2 can highly improve the system robustness with respect to model
Particle swarm optimization uncertainties.
Fractional order PID © 2014 Karabuk University. Production and hosting by Elsevier B.V. This is an open access article under
the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

1. Introduction swarm optimization (CRPSO) [5], Reinforcement Learning Autom-

ata (RLA) [6], Artificial Bee Colony (ABC) [7], Differential Evolution
In power generation systems, automatic voltage regulator (AVR) Algorithm (DEA) [8], Many Optimizing Liaisons (MOL) [9], Local
is utilized to maintain the terminal voltage of a synchronous Unimodal Sampling (LUS) [10], and Chaotic Ant Swarm (CAS) [11].
generator at a specified level. The AVR controls the consistency of CAS is a new search algorithm inspired by the biological behavior of
the terminal voltage by varying the exciter voltage of the generator ants in nature proposed by Li et al. [12]. However, it is a deter-
[1]. Due to the high inductance of the generator field windings and ministic process different from the conventional ant algorithm [13].
load variation, stable and fast response of the regulator is difficult to It combines the chaotic behavior of individual ants with the intel-
achieve. Therefore, it is important to improve the AVR performance ligent optimization action of an ant colony and thus it integrates the
and ensure stable and efficient response to transient changes in advantages of chaotic search and the powerful ability of swarm
terminal voltage. Various control structures have been proposed for collectiveness. Based on CAS algorithm, Li et al. developed a model
the AVR system, however, among these controllers the proportional which can be used to describe how an ant colony organizes itself to
plus integral plus derivative (PID) was the most preferable find the optimal path between a food source and the nest [14]. The
controller. The PID controller is distinguished by its robust perfor- CAS algorithm shows a great potential in solving difficult optimi-
mance over a wide range of operating conditions and simplicity of zation problems encountered in various fields such as parameter
structure design [2]. The design of the PID controller involves the identification of dynamic systems [13], fuzzy system identification
determination of three parameters which are the proportional, [15], and parameters tuning of PID controller [11].
integral, and derivative gains. In recent years, many intelligent Recently, large and growing body of literature has investigated
optimization algorithms were proposed to tune the PID gains of the the concept of fractional calculus in many control applications to
AVR system. Such algorithms include Particle Swarm Optimization enhance the performance of PID controller [16e18]. Fractional or-
(PSO) [3], Genetic algorithm (GA) [3,4], Craziness based particle der PID (FOPID) controller was first proposed by Podlubny in 1999
[19]. FOPID is a generalization of the PID in which the orders of
derivatives and integrals are non-integer [20]. The application of
FOPID controller was also employed to control AVR system [21e23].
E-mail address: [email protected].
Compared to conventional PID, FOPID can ensure good control
Peer review under responsibility of Karabuk University.

http://dx.doi.org/10.1016/j.jestch.2014.11.006
2215-0986/© 2014 Karabuk University. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/3.0/).
 M.A. Sahib / Engineering Science and Technology, an International Journal 18 (2015) 194e206 195

Table 1 Step Response

1.6
Transfer functions of the AVR system components. G_AVR
1.4 G_AVRm
AVR component Transfer function Range of the gain K Range of the time
constant T (s) 1.2
Amplifier Ga ¼ TaKsþ1
a 10e40 0.02e0.1
1

Amplitude
Exciter Ge ¼ TeKsþ1
e 1e10 0.4e1.0
0.8
Generator K 0.7e1 1.0e2.0
Gg ¼ Tg sþ1
g

0.6
Sensor Hs ¼ Ts Ksþ1
s 0.9e1.1 0.001e0.06
0.4

0.2
performance and improve the system robustness with respect to 0
model uncertainties [24]. However, due to fractional order in the 0 2 4 6 8 10 12
Time (seconds)
differentiator and integrator, realization of FOPID is performed with
high order discrete time controllers affecting the computational Fig. 2. Step response of the AVR system without PID controller.
load and memory size of the control algorithm. Therefore, various
approximation methods have been proposed to reduce the con-
troller's order. However, the so-called long memory principle
feature of the FOPID controllers will not be preserved after AVR system parameters considered in this work are; Ka ¼ 10.0,
approximation. Another property that is lost after approximation is Ta ¼ 0.1, Ke ¼ 1.0, Te ¼ 0.4, Kg¼1.0, Tg ¼ 1.0, Ks¼1.0, Ts ¼ 0.01
the optimality of controller [25]. [3,7,9,24,26]. With these parameter values the closed loop transfer
The main contribution of this paper is to propose a novel four function of the AVR system becomes:
term structure PID plus second order derivative (PIDD2) controller
for AVR system. The four gains of the PIDD2 are tuned by PSO al- DVt ðsÞ 0:1s þ 10
GAVR ¼ ¼
gorithm. The performance of the proposed PIDD2 is compared with DVref ðsÞ 0:0004s4 þ 0:045s3 þ 0:555s2 þ 1:51s þ 11
some PID controllers tuned by recently published modern heuristic 250ðs þ 100Þ
optimization algorithms such as MOL, GA, ABC, DEA, and LUS al- ¼

ðs þ 98:82Þðs þ 12:63Þ s2 þ 1:057s þ 22:04
gorithms. In addition, a comparison with the FOPID controller
tuned by CAS algorithm is also performed. The performance of the ~ 250
yG AVR ¼

proposed PIDD2 is further investigated using frequency response, ðs þ 12:63Þ s2 þ 1:057s þ 22:04
zero-pole map, and robustness analysis.
(1)
The remaining part of this paper is organized as follows. In
Section 2, the AVR system model is described. The AVR system with The transfer function of the AVR system (GAVR) have one zero at
PID controller is analyzed in Section 3. The proposed PIDD2 z ¼ 100, two real poles at s1 ¼ 98.82 and s2 ¼ 12.63, and two
controller is presented in Section 4. The PSO algorithm is explained complex poles at s3,4 ¼ 0.53 ± 4.66i. The GAVR can be approxi-
in section 5. In Section 6, the practical implementation issues of the mated by canceling the zero at 100 with the pole at 98.82 to
PIDD2 controller are addressed. Section 7 is devoted to computer obtain G~ ~
AVR . The unit step responses of GAVR and GAVR are shown in
simulation. Finally, Section 8 concludes the paper. Fig. 2. It can be observed from Fig. 2 that the AVR system GAVR
and its approximation G ~
AVR are almost similar and possess an
underdamped response with steady state amplitude value of 0.909,
2. AVR system model peak amplitude of 1.5 (Mp ¼ 65.43%) at tp ¼ 0.75, tr ¼ 0.42 s,
ts ¼ 6.97 s at which the response has settled to 98% of the steady
In synchronous generators, the AVR system is used to maintain state value.
the terminal voltage magnitude at a constant specified level. A
simple AVR system consists of four main components, namely
amplifier, exciter, generator, and sensor. Each component is modeled 3. Analysis of the AVR system with PID controller
by a first order system defined by a gain and a time constant. Table 1,
shows the four AVR main components transfer functions with their The response of the AVR can be improved by utilizing a
corresponding gain and time constants typical ranges [9]. controller in the forward path capable of processing the voltage
The arrangement of the AVR system components is shown in difference DVe(s) and producing a manipulated actuating signal.
Fig. 1. The terminal voltage DVt(s) of the generator is continuously Commonly, a PID controller is employed for this task due to its
sensed by the sensor and compared with the desired reference simple structure. The PID controller combines three control actions
voltage DVref(s). The difference between the reference and the related to the error signal in proportional, deferential, and integral
sensed terminal voltages (error voltage DVe(s)) is amplified through manners and its transfer function is given by:
the amplifier and used to excite the generator using the exciter. The
Ki
CPID ¼ Kp þ þ sKd (2)
s

where Kp, Ki, and Kd are the proportional, integral, and derivative
gains, Fig. 3 shows a block diagram of the AVR system with PID
controller. The general transfer function of the AVR system
controlled by a PID controller is given by

CPID Ga Ge Gg
GAVR_PID ¼ (3)
1 þ CPID Ga Ge Gg Hs
Fig. 1. AVR system block diagram.
196 M.A. Sahib / Engineering Science and Technology, an International Journal 18 (2015) 194e206

Substituting the transfer functions of the AVR system evaluating the optimum PID gains can be handled using an
components listed in Table 1 with their parameters and the transfer optimization problem in which an optimization algorithm is
function of the PID controller given by Equation (2) in (3) yields, employed. The optimization algorithm, such as PSO, uses an



0:1Kd s3 þ 0:1Kp þ 10Kd s2 þ 0:1Ki þ 10Kp s þ 10Ki
GAVR_PID ¼
 (4)
0:0004s5 þ 0:0454s4 þ 0:555s3 þ ð1:51 þ 10Kd Þs2 þ 1 þ 10Kp s þ 10Ki

The effect of the PID gain parameters on the overall AVR system objective or cost function to tune the PID gains. For example,
can be analyzed by plotting the closed loop zero-pole locus as a Panda et al., proposed the simplified PSO algorithm to design a
function of the PID gains. The zero-pole locus can be obtained when PID controller for the AVR system [9]. By investigating the zero-
Kp, Ki, and Kd are varied within the closed ranges 1  Kp  Kp_max, pole map of the overall transfer function (the AVR system with
0  Ki  Ki_max, and 0  Kd  Kd_max respectively. The initial state of the designed PID), given by [9],
the zero-pole locus can be easily obtained by setting Kp ¼ 1, Ki ¼ 0,

0:01772s3 þ 1:831s2 þ 5:899s þ 4:189

GAVR ¼
PID 0:0004s þ 0:045s4 þ 0:555s3 þ 3:282s2 þ 6:857s þ 4:189
5
(5)
44:3ðs þ 100:03Þðs þ 2:26Þðs þ 1:05Þ ~ 46:52
¼
yG AVRPID ¼
s2 þ 9:84s þ 46:52 ;
ðs þ 100:49Þðs þ 2:11Þðs þ 1:06Þ s2 þ 9:84s þ 46:52

and Kd ¼ 0 in Equation (4) and as a result the transfer function of

the AVR system reduces to that given by Equation (1) (without PID one can observe that the objective of the PID controller is to
controller). compensate the effect of two poles in the AVR system
The characteristic of the transient response of the AVR system at s1 ¼ 2.11 and s2 ¼ 1.06, thus the overall transfer
is closely related to the location of the closed-loop poles. From function GAVRPID can be approximated to G~
AVRPID . Fig. 4, shows
the design viewpoint, the adjustment of the PID gains may move the step responses of G ~
and G .
AVRPID AVRPID
the closed-loop poles to a desired location. Hence, with the use of From Fig. 4, it is observable that the step response of the AVR
the zero-pole locus method, it is possible to determine the values system and its approximation has been improved when using an
of the PID gains that will make the damping ratio of the domi- optimal PID controller. This is evident through an improved values
nant closed-loop poles as prescribed. However, a multi-gain root- of rise time tr ¼ 0.343, settling time ts ¼ 0.516 sec, maximum
locus is not an easy way to obtain and difficult to illustrate and overshoot Mp ¼ 1.95%, and damping ratio z ¼ 0.72.
plot on the complex plane. Alternatively, the problem of From the above analysis, it can be concluded that the PID
controller attempts to compensate the effect of two poles of the AVR
system. When the PID controller gain parameters are optimized, the
overall transfer function is approximately reduced from fourth to a
simple second order system. However, in a second-order system, the
maximum overshoot and the rise time of the unit step response
conflict with each other. Therefore, the improvement of the AVR
system response achieved by the conventional PID controller is a
compromise between maximum overshoot and rise time.

Fig. 3. AVR system with PID controller. 4. PID plus second order derivative controller (PIDD2)

The closed loop transfer function of the AVR system with opti-
1.4 mized PID controller can be approximated by a standard form of a
GAVR PID
second-order system given by
1.2 GAVR PID (approximated)
~ u2n
1 G AVRPID ¼ (6)
s2 þ 2zun s þ u2n
Amplitude

0.8
where un is the undamped natural frequency. The proposed
0.6 method is to modify the structure of the conventional PID
0.4 controller such that it can reduce the overall transfer function to
produce a modified form of Equation (6) in which an additional
0.2 zero is added at s ¼ a, such that,
 
0 s
0 0.2 0.4 0.6 0.8 1 1.2 1.4 a þ 1 u2n s~
Time (seconds) Gz ¼ ¼ G ~
þG (7)
s2 þ 2zun s þ u2n a AVRPID AVRPID

Fig. 4. Step response of the AVR system with PID controller.

 M.A. Sahib / Engineering Science and Technology, an International Journal 18 (2015) 194e206 197

then the step response of the modified system become

1 1 s ~ ~

Yz ¼ Gz ¼ GAVRPID þ G AVRPID
s s a
   
s 1~ 1~
¼ GAVRPID þ GAVRPID (8)
a s s

~
where ð1=sG AVRPID Þ is the unit step response (Y) of the original
approximated transfer function ðG ~
AVRPID Þ, thus,

s 1 _
Yz ¼ Y þY or yz ðtÞ ¼ yðtÞ þ yðtÞ (9)
a a
This means that, the step response of the modified second order
system with a zero at s ¼ a is given by the step response of the
original system plus a scaled version of its derivative. As the zero ~
Fig. 6. Effect of adding a zero to G AVRPID on tr, ts, Mp, and tp.

moves further to the left side of the complex plane (a increases),

the contribution of the derivative term yðtÞ _ decreases and the step
yðtÞ ¼ 1  eun t  un teun t (10)
response of the modified system starts to resemble the response of
the original approximated system. Conversely, as the zero moves The peak time tpz at which the maximum overshoot Mpz of the
closer to the origin from the left side (a decreases), the contribution modified response yz(t) occurs, can be found by substituting
of the derivative term yðtÞ _ increases resulting in an increased Equation (10) in (9), taking the derivative of yz(t), and equating to
overshoot, decreased rise and peak times (the step response be- zero yields,
comes faster). Fig. 5, shows the effect of adding a zero to the
 
approximated system defined by Equation (5) on the unit step 1 un  u2n
_ y_z ðtÞ ¼ y _
€ðtÞ þ yðtÞ ¼ eun t u2n t 1  þ ¼0 (11)
response. The scaling factor (1/a) of the derivative term yðtÞ is a a a
varied from 0 to 0.3.
From Fig. 5, it can be observed that when the contribution of the Solving equation (11) for t to get,
derivative term increases (1/a increases) the response becomes
1
faster (tr decrease) and possess higher overshoot peak (Mp in- t ¼ tpz ¼ ∞ or t ¼ tpz ¼ ; for a < un (12)
ðun  aÞ
crease). These results have been also illustrated in Fig. 6 where the
step response parameters (tr, ts, Mp, and tp) are plotted against the From Equation (12) choosing a value of a less than un (a < un)
same range of variation of the scaling parameter (1/a). will make the step response posses an overshoot given by,
In Fig. 6, the rise time is recorded as the time in which the
 
response takes to rise from 0 to 80% of the steady-state value. When
 1
Mpz ¼ yz tpz ¼ yz (13)
(1/a) increases, tr, ts, and tp decrease against an increase of Mp. un  a
When adding a zero to a second order system with under-
damped case (z < 1), such as the approximated system defined by
On the other hand, choosing a  un will positively eliminate the
Equation (5), the modified system will possess a faster response
overshoot.
versus an undesirable increase of Mp. Within the time interval
In the overdamped case (z > 1), where the poles are both real
tr  t  tp, the value of the original step response is 1  y(t)  1 þ Mp
_ located at s1 ¼ r1 and s2 ¼ r2, where r2 > r1, the unit step
and thus the value of its derivative is ε  ð1=aÞ yðtÞ  0 , where ε is a
response is given by [27],
positive real number. The value of ε depends proportionally on the
scaling parameter (1/a). Therefore the value of the modified
response is 1 þ ε  fyz ðtÞ ¼ yðtÞ þ ð1=aÞ yðtÞg_  1 þ Mp , and thus
Mpz will become greater than Mp, as well as tpz < tp, and trz < tr r2 er1 t  r1 er2 t
yðtÞ ¼ 1  (14)
where Mpz, tpz, and trz are the maxim overshoot, peak time, and rise r2  r1
time of the modified response yz(t).
In the critical damped case (z ¼ 1), where the poles are both Similarly, the peak time tpz at which the maximum overshoot
located at s ¼ un, the unit step response is given by [27], Mpz of the modified response yz(t) occurs can be found as in the
critical damped case to get,

1  r   r 
y_z ðtÞ ¼ y _
€ðtÞ þ yðtÞ ¼ er1 t 1  1  er2 t 1  2 ¼ 0 (15)
a a a

Solving Equation (15) for t to get,

 
1 a  r2
t ¼ tpz ¼ ln (16)
ðr2  r1 Þ a  r1

From Equation (16) choosing a value of a such that

(a < r1 < r2) will make the step response posses an overshoot.
Otherwise, choosing (r1 < a < r2) or (r1 < r2 < a) will eliminate
the overshoot.
From the previous analysis, in AVR system controller design, two
~
Fig. 5. Effect of adding a zero to G objectives are considered. The first objective is to modify the PID
AVR PID on the unit step response.
198 M.A. Sahib / Engineering Science and Technology, an International Journal 18 (2015) 194e206

controller such that, when optimized, the overall transfer function where i ¼ 1, 2,…, L, and L is the number of population (swarm size);
of the AVR system can be reduced to have the form defined by w is the inertia weight, c1 and c2 are two positive constants, called
Equation (7). The second objective is to direct the optimization the cognitive and social parameters respectively; ri1 and ri2 are
algorithm used to tune the controller parameters to minimize Mpz, random numbers uniformly distributed within the range [0, 1].
tpz, and trz as well as the settling time tsz. To achieve these objectives Equation (19) above is used to find the new velocity for the ith
a four term control type structure is proposed consisting of pro- particle, while Equation (20) is used to update the ith position by
portional, integral, derivative, and second order derivative terms adding the new velocity obtained by Equation (19).
(PIDD2) defined by, A simplified version of PSO (SSO) called “social only” suggested
by Kennedy is implemented by eliminating the personal influence
Ki (cognitive) term in the velocity update equation [31]. This can be
CPIDD2 ¼ Kp þ þ Kd s þ Kd2 s2 : (17)
s achieved by setting c1 in Equation (19) to zero, thus it becomes:

The difference between the proposed PIDD2 and the conven-  

tional PID controllers is the extra second order derivative term Vikþ1 ¼ wVikþ1 þ c2 ri2
k
Pgk  Xik (21)
added in the PIDD2 controller. This term is determined by the gain
parameter Kd2. Substituting the transfer functions of the AVR sys- The simplified PSO is also called Many Optimizing Liaisons
tem components listed in Table 1 with their parameters and the (MOL) to make it easy to distinguish from the original PSO [9]. MOL
transfer function of the proposed PIDD2 controller given by Equa- differs from PSO in that it eliminates the particle's best-known
tion (17) in (3) yields, position thus making the algorithm simpler.



0:1Kd2 s4 þ ð0:1Kd þ 10Kd2 Þs3 þ 10Kd þ 0:1Kp s2 þ 0:1Ki þ 10Kp s þ 10Ki
GAVR_PIDD2 ¼
 (18)
0:0004s5 þ 0:0454s4 þ ð10Kd2 þ 0:555Þs3 þ ð10Kd þ 1:51Þs2 þ 10Kp þ 1 s þ 10Ki

The proposed PIDD2 controller is expected to compensate the 6. PIDD2 implementation issues
effect of two AVR system poles, and hence reducing the overall
transfer function to that defined by Equation (7). With the new Presently, almost all control strategies are implemented as
proposed controller structure, the optimization algorithm digital algorithms in microprocessor-based equipment such as
employed for designing the PIDD2 controller will attempt to tune programmable logic controllers (PLCs) and digital signal processors
four gain parameters. (DSPs). To become applicable in such equipment, the PID control
algorithm has to be discretized using discretization methods. These
5. Particle swarm optimization methods can be applied similarly to discretize the proposed PIDD2
controller. The continuous time expression of the proposed PIDD2
The PSO algorithm is considered to be one of the most promising controller in ideal form is given by:
optimization techniques due to its simplicity, robustness, fast
convergence, and ease of implementation [28]. Solving optimiza- Zt

deðtÞ d2 e t
tion problem with PSO is based on the concept of social interaction uðtÞ ¼ Kp eðtÞ þ Ki eðtÞdt þ Kd þ Kd2 : (22)
in which a population of individual solutions called particles is
dt dt 2
0
employed for the searching process [29]. The particles are grouped
in a finite set called swarm and are updated iteratively. In each Applying the trapezoidal approximation to discretize the inte-
iteration, the particles exchange previously discovered information gral term and the backward finite differences approximation to
with neighbors and use these information to update their new discretize the first and second derivative terms [32] in Equation
position. The new positions of particles are calculated by adding (22) to get an approximated discrete transfer function of the PIDD2
their previous position to their corresponding updated velocity given by,
values. In PSO algorithm, updating the velocity for each particle is
  
the most important step. The velocity is updated using the previous UðzÞ K Ts z þ 1 z1 z1 2
velocity (inertia), personal influence (cognitive), and social influ- ¼ CPIDD2 ðzÞ ¼ Kp þ i þ Kd þ Kd2
EðzÞ 2 z1 Ts z Ts z
ence (social) components. The inertia component prompts the
(23)
particle to move in the same previous direction and velocity. The
cognitive component improves the new particle's position by
where Ts is the sampling interval. The common practical imple-
comparing it with the best previous position found associated with
mentation problems of the PID controller are the integral windup
this particle. The social component makes the particle follow the
and derivative kick problems. Remedies for the integral windup
best neighbor's direction. The modified velocity and position of
problem used with PID implementation can also be applied for the
each particle are calculated according to the following equations
PIDD2 controller. However, due to the second derivative term of the
[30]:
proposed PIDD2 controller, the derivative kick problem becomes a
    major concern in practical implementation. A drawback with the
Vikþ1 ¼ wVikþ1 þ c1 ri1
k
Pik  Xik þ c2 ri2
k
Pgk  Xik (19) first order derivative term is that it will amplify the input signal
with a gain directly related to its frequency (linear increasing
Xikþ1 ¼ Xik þ Vikþ1 (20) magnitude Bode plot with 20 dB per decade). The effect of this
drawback will be doubled with the second order derivative term
and the gain become directly related to the square of its frequency
 M.A. Sahib / Engineering Science and Technology, an International Journal 18 (2015) 194e206 199

(linear increasing magnitude Bode plot with 40 dB per decade). The

amplification effect is more evident when the error signal exhibit 2 3 2 32
high frequency components caused by measurement noise, load
K 6 s 7 6 7
disturbance, and/or set point changes. For example, when an ~
C ¼ Kp þ i þ Kd 6 7 þ Kd2 6 s 7 ; (24)
PIDD2 s 4 sTd 5 4 sTd2 5
abrupt (stepwise) change of the set-point value occurs, the first and 1þ N1 1þ N2
second derivative actions will be very large and this results in an
undesirable spike (first plus second derivative kick) in the control
variable signal. As a result, the actuator unit will experience a where Td and Td2 are the first and second derivative time constants
rapidly changing command signal that could be detrimental to the respectively. The filters coefficients N1 and N2 can be adjusted to set
operation of the unit. This problem can be solved by limiting the the cutoff frequencies of the first and second order derivative filters
bandwidth of the first and second order derivative actions with a respectively. When N1 and N2 approach infinite, Equation (24) re-
first and second order low-pass filters respectively. In this context, duces to the ideal form CPIDD2 . The high-frequency gains of the
the PIDD2 controller defined by Equation (17) can be modified to be modified first and second derivative terms are

Fig. 7. Pseudocode for NMF used to realize first and second order derivative actions.
200 M.A. Sahib / Engineering Science and Technology, an International Journal 18 (2015) 194e206

2 3 2 32 performance criterion [34]. The simulation parameters of the PSO

algorithm are listed in Table 2.
6 s 7 6 s 7
lim Kd 6 7 ¼ Kd ðN1 =Td Þ and lim Kd2 6 7 The searching range of the PIDD2 gains and their corresponding
4 5 4 5
s/∞ 1 þ sT
N1
d s/∞ 1 þ sTNd2
2
velocity constraints are defined in Table 3.
To improve the search process of any optimization algorithm, it
¼ Kd2 ðN2 =Td2 Þ2 is necessary to bound the dimensions of the searching space. In PID
controller tuning, defining the maximum limits of the gains is
(25)
important for control system stability. From recent literature re-
2
With the modified PIDD controller defined by Equation (24), sults, it has been found that optimum PID gain values used to
the optimization algorithm can also be modified to tune the filters control the AVR system GAVR are within [0, 1.5], [0, 1], [0, 1] for Kp, Ki,
coefficients N1 and N2 along with the four gain parameters. In this and Kd respectively [3,9,10]. However, for the proposed PIDD2 the
case, the optimization objective is to minimize tr, ts, Mp, and to search ranges of all gains are expanded to be [0.0001, 3]. The
minimize the maximum range of the controller output. maximum and minimum velocity limits determine the resolution,
An alternative method for smoothing the first and second de- or fitness, with which regions be searched between the present
rivative actions is to use a nonlinear median filter (NMF) [33], PIDD2 gain value and the target value. If these limits are chosen
which is widely applied in image processing. The NMF compares high, the PIDD2 gain values may move erratically, going beyond a
several data points around the current point and selects their good solution. On the contrary, if the limits are chosen too small the
median for the control action. As a result, the high frequency gains may not explore sufficiently beyond local solutions. An
components (unwanted spikes) resulting from a step command, effective velocity limit value is chosen to be 20% of the corre-
noise, or disturbance are removed completely. Fig. 7 illustrates the sponding maximum gain value [35].
pseudocode of the NMF for the first and second derivative actions. For each particle (set of PIDD2 gains), the closed-loop system
Unlike lowpass filters, which averages past values, NMF is stability is tested using the “isstable” Matlab function. If the func-
capable of removing extraordinary derivative values resulting from tion returns a logical true value, then the solution is feasible and its
sudden changes in the error signal. Fig. 8 shows an example of an fitness value is considered. Otherwise, if the function returns a
error signal, e(t), having high frequency components (abrupt logical false, then the closed-loop system is unstable and hence the
changes and sharp edges). The first and second derivatives of e(t) solution is infeasible. Infeasible solutions are excluded by penal-
are computed using NMF. izing them with very large fitness value.
The error signal example, shown in Fig. 8, has abrupt changes at
time instants 0.5, 3.5, 4, and 4.9 s. When a backward difference
method is used to approximate the first and second order de- 7.1. Transient response analysis
rivatives, unwanted spikes will occur at these instants. However,
with NMF the undesired spikes are completely removed and thus The transient response of the proposed PIDD2 controller tuned
resulting in a nonaggressive control signal. with PSO is analyzed by comparing the unit-step response with
different PID controllers. The PID controllers were designed in
7. Simulation results and discussion recent literature using PSO [3], MOL [9], GA [3], ABC [7], DEA [7],
and LUS [10] for the same AVR system. Fig. 9, shows a comparison of
In this section, the proposed PIDD2 controller is tested in con- the AVR terminal voltage step response of the proposed PIDD2 and
trolling the AVR system GAVR defined by Equation (1). The perfor- different PID controllers. Each PID controller is associated with one
mance of the PIDD2 controller is compared with conventional PID of the aforementioned tuning algorithms and one objective func-
controllers tuned by recently published modern heuristic optimi- tion. The different objective functions used are the ITAE, integral of
zation techniques. The PIDD2 is also compared with FOPID time multiplied by squared error (ITSE) [7], f function [3], and OF4
controller. In addition, transient response, zero-pole, frequency function [10], defined by,
response, and robustness analysis are performed on the proposed
PIDD2 controller. The realization of the proposed PIDD2 controller
and its discrete implementation is also tested in SIMULINK®. The Table 2
PSO algorithm is employed to tune the PIDD2 gain parameters using PSO searching parameters.
the integral of time multiplied by absolute error (ITAE)
Parameter Value

Number of iterations (N) 50

4 Number of trials (T) 10
x(t)
dx(t)/dt (first order NMF) Swarm size (L) 30
3 Acceleration constants (c1 ¼ c2) 2
d x(t)/dt (second order NMF)
Inertia weight factor (w) [0.9:0.014:0.2]
2

Table 3
e(t)

0
Searching range of parameters.
-1
Parameter Min. value Max. value
-2 Kp 0.0001 3
Ki 0.0001 3
-3
Kd 0.0001 3
Kd2 0.0001 3
-4
0 1 2 3 4 5 6 7 8 vK p 0.6 0.6
time (sec)
vK i 0.6 0.6
vK d 0.6 0.6
Fig. 8. An example illustrating computation of first and second order derivatives using
vKd2 0.6 0.6
NMF.
 M.A. Sahib / Engineering Science and Technology, an International Journal 18 (2015) 194e206 201

1.4 Table 5
CAS algorithm parameters [24].
1.2
Parameter Value
1 Number of ants (K) 20
PIDD2 / PSO / ITAE Positive constants (a, b) (300, 2/3)
Amplitude

0.8 PID / PSO / ITAE Organization factor of ant i (ri) 0.04 þ 0.1  rand( )
PID / MOL / ITAE [9] Initial state of ant i (yi(0)) 0.999
0.6 PID / GA / f [3] jd ðd ¼ 1; 2; …; 5Þ 7:5=ud
PID / PSO / f [3] Number of iterations 300
0.4 PID / ABC / ITSE [7]
PID / DEA / ITSE [7] rand( ) is a uniformly distributed number in [0, 1].
0.2 PID / PSO / ITSE [7] ud is the interval of search of the d-th controller parameter.
PID / LUS / OF4 [10]
0
0 0.5 1 1.5 2 2.5 3
Time (sec)
Also, the PSO-PIDD2 is compared with PSO-PID [3] and PSO-
Fig. 9. Terminal voltage step response of the AVR system with different controllers. FOPID controllers [24]. The reciprocal of f defined in Equation
(28) is considered as the objective function to tune the PSO-PID,
PSO-FOPID, CAS-PID, and CAS-FOPID with two cases; b ¼ 1 and
Ztss b ¼ 1.5. The terminal voltage step responses of the AVR system
ITAEmin ¼ tjeðtÞjdt; (26) controlled by PSO-PIDD2, PSO-PID, PSO-FOPID, CAS-PID and CAS-
0 FOPID controllers are shown in Fig. 10 with b ¼ 1 and b ¼ 1.5.
As can be seen from Fig. 10, the response of the PIDD2 is much
Ztss better than the PID and FOPID controllers tuned with PSO and CAS
ITSEmin ¼ te2 ðtÞdt; (27) algorithms in both cases (i.e. b ¼ 1 and b ¼ 1.5). This can be clearly
0 observed from the time performance indices of all controllers listed in
Table 6.
1
fmax ¼

 ; (28) It is observed from Table 6 that the PSO-PIDD2 has the best
1  eb Mp þ Ess þ eb ðts  tr Þ performance compared to PSO/CAS-PID and PSO/CAS-FOPID con-
Ztss trollers. The terminal voltage step response of the AVR system
OF4min ¼ 0:8* e2 ðtÞdt þ 0:1*ts þ 0:1*Mp ; (29) controlled by the proposed PIDD2 controller has the smallest values
of Mp, Ess, tr, and ts highlighted in bold.
0
The transfer function of the FOPID controllers defined by the
respectively. In Equations (26)e(29), tss is the time at which the
parameters listed in Table 6, are then implemented with integer
response reaches steady state, b is a weighting factor, and Ess is the
orders transfer function using Oustaloup recursive distribution of
steady state error.
poles and zeroes approximation [36]. The integer orders transfer
From Fig. 9, it can be observed that the proposed PIDD2 possess
function obtained by Oustaloup approximation will have an order
a superior step response behavior compared to other PID con-
equal to 12 [24]. This fact adds another preference to the PIDD2
trollers. Table 4, lists the numerical results of the response com-
related to implementation complexity. It is worth noting that, the
parison including; controller parameters, the time domain
proposed PIDD2 controller can be extended to fractional order
performance indices (Mp, tr, ts, and tp), and the objective function
PIDD2 (PIl Dm Dm2 ) where l, m, and m2 are non-integer (fractional)
values.
orders of the integral, first and second order derivatives parts
It is clear from Table 4, that the best response performance
respectively. The complexity of this controller is evident due to
indices values, highlighted in bold, are those obtained with the
the increase in the number of control parameters. There are
proposed PIDD2 controller (Mp ¼ 0, tr ¼ 0.0929, ts ¼ 0.1635, and
seven different parameters (Kp, Ki, Kd, Kd2, l, m, and m2) that have
tp ¼ 0.32). Therefore, in comparison to all PID controllers, the PIDD2
to be tuned. The challenge of this work is to develop a realizable
has the ability to achieve the fastest (minimum tr and ts), most
FOPIDD2 controller that exhibits a robust performance with
accurate (minimum response oscillation), and most stable (mini-
fewer parameters, yet achieving the same design requirements.
mum overshoot) unit step response.
The key point is to look for acceptable and realizable approxi-
The proposed PIDD2 controller designed by PSO is compared
mations of sl, sm, and sm2 which is recommended for future
with PID and FOPID controllers designed using CAS algorithm
investigation.
[12,13,24]. The CAS algorithm is implemented using the parameters
listed in Table 5.

Table 4
Controller parameters and response performance indices of different controllers.

Controller/algorithm/OF Controller parameters Mp % tr ts tp Obj. value

0.1 / 0.9 ±2%
Kp Ki Kd Kd2

PIDD2/PSO/ITAE 2.7784 1.8521 0.9997 0.07394 0 0.0929 0.1635 0.3200 0.0018

PID/PSO/ITAE 1.3541 0.9266 0.4378 - 18.805 0.1493 0.8146 0.3276 0.0329
PID/MOL/ITAE [9] 0.5857 0.4189 0.1772 e 1.9539 0.3433 0.5155 0.7036 0.0464
PID/GA/f [3] 0.8861 0.7984 0.3158 e 8.6532 0.2041 0.6058 0.4222 1.1982
PID/PSO/f [3] 0.6568 0.5393 0.2458 e 1.1652 0.2722 0.4111 1.9200 1.4480
PID/ABC/ITSE [7] 1.6524 0.4083 0.3654 e 25.035 0.1559 3.0939 0.3629 0.0177
PID/DEA/ITSE [7] 1.9499 0.4430 0.3427 e 32.830 0.1513 2.6494 0.3636 0.0220
PID/PSO/ITSE [7] 1.7774 0.3827 0.3184 e 30.048 0.1609 3.3994 0.3909 0.0238
PID/LUS/OF4 [10] 0.6190 0.4222 0.2058 e 0.5900 0.3123 0.4778 0.6008 0.1677
202 M.A. Sahib / Engineering Science and Technology, an International Journal 18 (2015) 194e206

1.2
PSO-PIDD2
1 PSO-PID
CAS-PID
PSO-FOPID
0.8
CAS-FOPID
Amplitude

0.6 1

0.4 0.9

0.2 0.8
0 0.2 0.4 0.6 0.8 1

0
0 1 2 3 4 5
Time (sec)
(a) Fig. 11. Zero-pole map of the AVR system controlled by PIDD2.
1.2

1  
1845:2 s þ1
100
~
G ¼ (31)
0.8 AVR_PIDD2
PSO-PIDD2 ðs þ 75:53Þ ðs þ 24:43Þ
Amplitude

PSO-PID
0.6 1 Comparing Equation (31) with the overdamped case of Equation
CAS-PID
PSO-FOPID (7) yields, a ¼ 100, r1 ¼ 24.43, and r2 ¼ 75.53 with r1 < r2 < a. In this
0.4 0.9 CAS-FOPID case, the system response possess no overshoot and this can be
ensured by substituting the values of a, r1, and r2 in Equation (16).
0.2 0.8 The ration of (a  r2)/(a  r1) inside the logarithm function is less
0 0.2 0.4 0.6 0.8 1
than one, thus resulting in a negative time value which indicates no
0
0 1 2 3 4 5 overshoot exists in the system's response.
Time (sec)
(b)

Fig. 10. Step response of AVR system controlled by PSO-PIDD2, PSO-PID, CAS-PID, PSO-
7.3. Frequency response analysis
FOPID, and CAS-FOPID (a) b ¼ 1 (b) b ¼ 1.5.
The frequency response of the AVR system with the proposed
PIDD2 controller is analyzed. The magnitude and phase plots of the
7.2. Zero-pole analysis AVR with PIDD2 controller is shown in Fig. 12. The peak gain, phase
margin, delay margin and bandwidth obtained from the system's
The overall closed-loop transfer function of the AVR system with frequency response are depicted in Table 7 and compared with
the proposed PIDD2 controller is of 5th order given by different controllers.

18:4855ðs þ 100Þðs þ 10:02Þðs þ 2:501Þðs þ 0:9994Þ

GAVR_PIDD2 ¼ (30)
ðs þ 75:53Þ ðs þ 24:43Þ ðs þ 10:04Þ ðs þ 2:502Þ ðs þ 0:9994Þ

The zero-pole map of the AVR system with the proposed PIDD2 As shown in Table 7, the PIDD2 is the most stable system
controller is shown in Fig. 11. compared to other controllers. The AVR with PIDD2 controller have
It can be observed that the system possess three zero-pole minimum peak gain 0 dB at 0 Hz, maximum phase margin 180 ,
cancellation pairs located at 1, 2.5, and 10, two real domi- infinite delay margin (smallest time delay required to make the
nant poles at s1 ¼ 24.43 and s1 ¼ 75.53, and one real zero at system unstable), and maximum bandwidth (fastest response). It is
z1 ¼ a ¼ 100. Due to the three zero-pole cancellation, the overall worth noting that, a wide bandwidth allows the system to follow
transfer function in Equation (30) can be approximated to be, arbitrary inputs accurately.

Table 6
Controller parameters and performance indices of PSO-PIDD2, PSO-PID, CAS-PID, PSO-FOPID, and CAS-FOPID.

Algorithm-controller Controller parameters Mp % Ess tr ts

0.1 / 0.9 ±2%
Kp Ki Kd Kd2 m l
PSO-PIDD2 2.778 1.852 0.999 0.074 - - 0 1.06e08 0.0929 0.1635
PSO-PID (b ¼ 1) 0.6570 0.5389 0.2458 e e e 1.1601 1.63e-05 0.2721 0.4110
CAS-PID (b ¼ 1) 0.6746 0.6009 0.2618 e e e 1.7686 1.04e-05 0.2574 0.3856
PSO-FOPID (b ¼ 1) 1.6264 0.2956 0.3226 e 1.1980 1.3183 5.4124 0.009037 0.1567 2.6848
CAS-FOPID (b ¼ 1) 1.0537 0.4418 0.2510 e 1.1122 1.0624 3.8524 0.001733 0.2191 0.5372
PSO-PID (b ¼ 1.5) 0.6254 0.4577 0.2187 e e e 0.4394 4.68e-06 0.3003 0.4606
CAS-PID (b ¼ 1.5) 0.6202 0.4531 0.2152 e e e 0.4026 5.30e-06 0.3045 0.4676
PSO-FOPID (b ¼ 1.5) 1.6986 0.1797 0.3122 e 1.2081 1.8373 5.7732 0.043639 0.1579 33.518
CAS-FOPID (b ¼ 1.5) 0.9315 0.4776 0.2536 e 1.0838 1.0275 2.8362 7.18e-04 0.2297 0.8949
 M.A. Sahib / Engineering Science and Technology, an International Journal 18 (2015) 194e206 203

Bode Diagram
0
Magnitude (dB)

-10

-20

-30

0
Phase (deg)

-45

-90
10 10 10 10 10
Frequency (rad/s)
Fig. 14. Step response curves ranging from 50% to þ50% for Te.
Fig. 12. Bode diagrams of the AVR controlled by PIDD2.

Table 7
Bode analysis of different AVR controllers.

Controller/algorithm/OF Peak gain dB Phase margin Delay Bandwidth

(deg.) margin

PIDD2/PSO/ITAE 0 (0 Hz) 180 Inf. 23.5031

PID/PSO/ITAE 1.79 (1.32 Hz) 79.3 0.1207 13.9142
PID/MOL/ITAE [9] 0 (0 Hz) 180 Inf. 6.3391
PID/GA/f [3] 0.17 (0.17 Hz) 116.2 0.2926 10.6614
PID/PSO/f [3] 0.07 (0.11 Hz) 166.9 2.6147 8.3137
PID/ABC/ITSE [7] 2.87 (1.21 Hz) 69.4 0.1109 12.8798
PID/DEA/ITSE [7] 4.20 (1.23 Hz) 58.4 0.0916 12.8006
PID/PSO/ITSE [7] 3.76 (1.16 Hz) 62.2 0.1033 12.1825
PID/LUS/OF4 [10] 0 (0 Hz) 180 Inf. 7.1673
CAS-FOPID (b ¼ 1) [24] 0.0053 (0.01 Hz) 178.5 23.324 9.9543 Fig. 15. Step response curves ranging from 50% to þ50% for Tg.
CAS-FOPID (b ¼ 1.5) [24] 0.0003 (0.01 Hz) 179.2 39.151 9.6367

7.4. Robustness analysis

Robustness analysis is used to evaluate the controller ability to

tolerate uncertainties exists in some system parameters. In this
subsection, the PIDD2 controller is tested against uncertainties of
AVR system parameters. The uncertainties of the AVR model are
specified in terms of variations in the amplifier, exciter generator,
and sensor time constants (Ta, Te, Tg, and Ts respectively) above and
below their nominal values. The variation range of the time con-
stants is chosen to be ±50% of their nominal values with a 25% step
size. Figs. 13e16 show step responses of the PIDD2 controlled AVR
system with Ta, Te, Tg, and Ts time constants variations about
nominal responses respectively. Fig. 16. Step response curves ranging from 50% to þ50% for Ts.
It can be realized from Figs. 13e16, that the deviations of
response curves (±50% and ±25%) from the nominal response for
the selected time constant parameters are within a small range. perform properly despite such large variations. Tables 8 and 9
This can ensures the ability of the PIDD2 to maintain stability and to present a summary of the PIDD2 robustness analysis results and
list the total deviation ranges and maximum deviation percentage
of the system respectively.
From Table 9, the average deviation of maximum overshoot,
settling time, rise time and peak time are 3%, 138%, 41% and 188%
respectively. The ranges of total deviation are acceptable and are
within limit. Therefore, it can be concluded that the AVR system
with the proposed PIDD2 controller is robust and can still perform
acceptable control behavior.

7.5. Digital implementation

The realization of the proposed PIDD2 controller and its discrete

implementation is tested in SIMULINK® and compared with PID/
MOL [9] and PID/GA [3] discrete controllers. The general Simulink
Fig. 13. Step response curves ranging from 50% to þ50% for Ta. model of the AVR control system is shown in Fig. 17.
204 M.A. Sahib / Engineering Science and Technology, an International Journal 18 (2015) 194e206

Table 8 0.008 and 0.0275 (Ts is set to 0.01). The response of the PIDD2,
Robustness analysis results of the AVR system with the proposed PIDD2 controller. PID/MOL [9], and PID/GA [3] controllers are tested at steady state by
Parameter Rate of change (%) Peak value (pu) ts tr tp subjecting a disturbance load signals of values equal to þ10%
Ta 50% 0.9994 0.3352 0.1226 0.7836
and 10% of the set point at times 3 and 5 s respectively. Figs. 18
25% 0.9967 0.2674 0.0897 0.4794 and 19 show the set point responses due to the unit step input at
þ25% 1.0243 0.2964 0.1005 0.2394 t ¼ 0 and responses due to load disturbances at t ¼ 3 and 5 s along
þ50% 1.0514 0.4038 0.1084 0.2476 with the controller outputs.
Te 50% 0.9985 0.5242 0.0395 1.6345
Compared to PID/MOL [9] and PID/GA [3] controllers, the pro-
25% 0.9972 0.2615 0.0685 1.0805
þ25% 1.0173 0.1770 0.1139 0.3379 posed PIDD2 with NMF (PIDD2/NMF) posses an improved set point
þ50% 1.0336 0.6386 0.1325 0.3698 and load disturbance responses as shown in Fig. 18. The responses
Tg 50% 0.9910 0.3119 0.0362 0.0717 of PIDD2 with filtered first and second derivative actions (PIDD2/
25% 0.9940 0.1261 0.0648 0.8705
N1N2) are faster than those of PID/MOL [9] and PID/GA [3] con-
þ25% 1.0096 0.1952 0.1189 0.4176
þ50% 1.0186 0.2243 0.1430 0.4687
trollers, however, it has the highest maximum overshoots values.
Ts 50% 0.9997 0.1896 0.1067 0.3801 In AVR control system, the controller actions are carried out as a
25% 0.9997 0.1774 0.1000 0.3532 response to load disturbance (regulating system) not to set point
þ25% 0.9996 0.1471 0.0857 0.2703 changes (tracking system). Therefore, in Fig. 19, only the responses
þ50% 1.0003 0.1285 0.0790 0.2441
to load disturbances are shown. It can be observed that the range of
the controller output signals for the PIDD2/NMF, PID/MOL [9], and
Table 9 PID/GA [3] controllers are ±3.8, ±0.7, and ±0.44 respectively.
Total deviation ranges and maximum deviation percentage of the system. However, for PIDD2/N1N2 controller, the range of the controller
Parameter Total deviation range/max deviation percentage (%) output signal exceeds ±4.
Peak value (pu) ts tr tp
Controllers are designed to work with nonlinear behavior of
0.9997 0.1635 0.0929 0.3200 process actuators. The actuator device, such as the amplifier in the
AVR system, has a limited range of input and output operation. Such
Ta 0.0547/5% 0.1364/147% 0.0329/32% 0.5442/145%
Te 0.0364/3% 0.4616/291% 0.0930/57% 1.2966/411% limitations appear at the input of the actuator and are modeled
Tg 0.0276/2% 0.1858/91% 0.1068/61% 0.7935/172% with a non-linear element having saturation characteristics.
Ts 0.0007/0% 0.0611/21% 0.0277/15% 0.1369/24% Moreover, when abrupt change occurs in the system output due to
Average 0.0299/3% 0.2112/138% 0.0651/41% 0.6928/188%
a load disturbance, the controller output will exhibit a large spike
values similar to those of the PIDD2/NMF shown in Fig. 19. These
spikes are mainly due to the first and second derivative actions and
The controller subsystem, shown in Fig. 17, is implemented by a could be detrimental to the operation of the actuator unit. To avoid
discrete PIDD2 or PID controller having specifications defined in subjecting the actuator unit to such large controller output values, a
Table 10. constrained action defined by the maximum and minimum output
The sampling time (Ts) is chosen according to the rule of thumb range limits of the actuator. However, in this case the integral action
suggested by Astrom and Wittenmark such that the product of Ts will produce an inaccurate and highly excessive value causing
and the gain crossover frequency (uc in radians per second) of the oscillation and slowing down the transient response. This behavior
loop gain ðCPIDD2 Ga Ge Gg Hs Þ, is between 0.15 and 0.5 [32]. The gain is called the integrator windup problem. This can be solved by
crossover frequency of the AVR control system loop gain is uc ¼ 18.2 several anti-windup algorithms such as the configuration sug-
radian per second. Thus, an appropriate sampling time is between gested by Wilkie et al. [37].

Fig. 17. Simulink model of the AVR control system.

 M.A. Sahib / Engineering Science and Technology, an International Journal 18 (2015) 194e206 205

Table 10
Controller subsystem specifications.

Controller Kp Ki Kd Kd2 Filter coefficient(s) Controller formula C(z)

Integration method Filter method Filter method
h i
PIDD2 2.7784 1.8521 0.9997 0.07394 e Kp þ K2i Ts zþ1
z1 þ Kd NMF 1 þ Kd2 NMF 2
Trapezoidal NMF1 NMF2 h i h i h i2
PIDD2 2.7784 1.8521 0.9997 0.07394 N1 ¼ 30 Kp þ K2i Ts zþ1
z1 þ Kd N1
N 1 Ts z þ Kd2 N2
N 2 Ts z
1þ z1 1þ z1
Trapezoidal Backward Euler Backward Euler N2 ¼ 80 h i h i
PID/MOL [9] 0.5857 0.4189 0.1772 e N ¼ 30 Kp þ K2i Ts zþ1
z1 þ Kd N
1þNT sz
z1
Trapezoidal Backward Euler
PID/GA [3] 0.8861 0.7984 0.3158 e N ¼ 30
Trapezoidal Backward Euler

1.4 8. Conclusion

1.3 1 In this paper, a novel PID plus second order derivative controller
(PIDD2) is proposed to control AVR system. The proposed PIDD2
Terminal Voltage (p.u.)

1.2
consists of four control terms; proportional, integral, derivative,
0.9
5 5.5 6 and second derivative. The PSO algorithm with the integral of time
1.1
multiplied by absolute error (ITAE) performance criterion is used to
tune the four gains of the PIDD2 controller. The performance of the
1
PIDD2/NMF
AVR with PIDD2 is compared with several PID controllers tuned by
PIDD2/N1N2 recently proposed approaches, such as MOL, GA, ABC, DEA, and LUS.
0.9
PID/MOL In addition, the proposed PIDD2 is compared with the FOPID
PID/GA controller designed by using CAS algorithm. Simulation results
0.8
0 1 2 3 4 5 6 7 8
Time (sec) show a superior response performance of the proposed PIDD2.
Moreover, the frequency response, zero-pole, and robustness
Fig. 18. Set-point and disturbance responses the AVR control system. analysis performed on the PIDD2 controller showed more robust
stability and better performance characteristics than the PID and
FOPID controllers.
4
PIDD2/NMF
References
3 PIDD2/N1N2
PID/MOL
2 [1] P. Kundur, N.J. Balu, M.G. Lauby, Power System Stability and Control, vol. 7,
PID/GA
McGraw-Hill, New York, 1994.
[2] A. Kiam Heong, G. Chong, L. Yun, PID control system analysis, design, and
Controller Output

1
technology, IEEE Trans. Control Syst. Technol. 13 (4) (2005) 559e576.
0 [3] G. Zwe-Lee, A particle swarm optimization approach for optimum design of
PID controller in AVR system, IEEE Trans. Energy Convers. 19 (2) (2004)
-1 384e391.
[4] P. Wang, D.P. Kwok, Optimal design of PID process controllers based on ge-
-2 netic algorithms, Control Eng. Pract. 2 (4) (1994) 641e648.
[5] V. Mukherjee, S.P. Ghoshal, Intelligent particle swarm optimized fuzzy PID
-3 controller for AVR system, Electr. Power Syst. Res. 77 (12) (2007) 1689e1698.
[6] M. Kashki, Y. Abdel-Magid, M. Abido, A reinforcement learning automata
-4 optimization approach for optimum tuning of PID controller in AVR system,
3 3.5 4 4.5 5 5.5 6
time (sec) in: D.-S. Huang, et al. (Eds.), Advanced Intelligent Computing Theories and
Applications. With Aspects of Artificial Intelligence, Springer Berlin, Heidel-
Fig. 19. Controller output. berg, 2008, pp. 684e692.
[7] H. Gozde, M.C. Taplamacioglu, Comparative performance analysis of artificial
bee colony algorithm for automatic voltage regulator (AVR) system, J. Franklin
Inst. 348 (8) (2011) 1927e1946.
[8] G. Reynoso-Meza, et al., Controller tuning using evolutionary multi-objective
The results as summarized in Table 11 indicate that the response optimisation: current trends and applications, Control Eng. Pract. 28 (0)
of the proposed PIDD2/NMF controller outperforms the responses (2014) 58e73.
of the PIDD2/N1N2, PID/MOL [9], and PID/GA [3] in terms of [9] S. Panda, B.K. Sahu, P.K. Mohanty, Design and performance analysis of PID
controller for an automatic voltage regulator system using simplified particle
maximum overshoot, rise time, and settling time. The best swarm optimization, J. Franklin Inst. 349 (8) (2012) 2609e2625.
response performance indices values of the proposed PIDD2/NMF [10] P.K. Mohanty, B.K. Sahu, S. Panda, Tuning and assessment of proportio-
controller are highlighted in bold. naleintegralederivative controller for an automatic voltage regulator system
employing local unimodal sampling algorithm, Electr. Power Compon. Syst. 42
(9) (2014) 959e969.
[11] H. Zhu, et al., CAS algorithm-based optimum design of PID controller in AVR
system, Chaos Solitons Fractals 42 (2) (2009) 792e800.
Table 11 [12] L. Li, et al., An optimization method inspired by “chaotic” ant behavior, Int. J.
Performance comparison. Bifurcation Chaos 16 (08) (2006) 2351e2364.
[13] L. Li, et al., Parameters identification of chaotic systems via chaotic ant swarm,
Controller Set-point response Load-disturbance response Chaos Solitons Fractals 28 (5) (2006) 1204e1211.
[14] L. Li, et al., Chaoseorder transition in foraging behavior of ants, Proc. Natl.
Mp % tr ts Mp % tr ts
Acad. Sci. 111 (23) (2014) 8392e8397.
PIDD2/NMF 10 0.08 0.16 9.3 0.09 0.18 [15] L. Li, Y. Yang, H. Peng, Fuzzy system identification via chaotic ant swarm,
PIDD2/N1N2 39 0.10 0.49 38 0.11 0.50 Chaos Solitons Fractals 41 (1) (2009) 401e409.
PID/MOL [9] 10 0.43 1.27 10 0.45 0.28 [16] S. Das, et al., A novel fractional order fuzzy PID controller and its optimal time
domain tuning based on integral performance indices, Eng. Appl. Artif. Intell.
PID/GA [3] 29 0.28 1.36 28 0.30 1.36
25 (2) (2012) 430e442.
206 M.A. Sahib / Engineering Science and Technology, an International Journal 18 (2015) 194e206

[17] A. Rajasekhar, R. Kumar Jatoth, A. Abraham, Design of intelligent PID/ [27] K. Ogata, Modern Control Engineering, Prentice Hall, 2010.
PIlDm speed controller for chopper fed DC motor drive using opposition [28] J. Kennedy, R. Eberhart, Particle swarm optimization, in: IEEE International
based artificial bee colony algorithm, Eng. Appl. Artif. Intell. 29 (0) (2014) Conference on Neural Networks, 1995.
13e32. [29] A. Ertas, Optimization of fiber-reinforced laminates for a maximum fatigue life
[18] R. El-Khazali, Fractional-order controller design, Comput. Math. Appl. 66 (5) by using the particle swarm optimization. Part II, Mech. Compos. Mater. 49 (1)
(2013) 639e646. (2013) 107e116.
[19] I. Podlubny, Fractional-order systems and PI/sup/spl lambda//D/sup/spl mu//- [30] S. Yuhui, R. Eberhart, A modified particle swarm optimizer, in: The 1998 IEEE
controllers, IEEE Trans. Auto. Control 44 (1) (1999) 208e214. International Conference on Evolutionary Computation Proceedings, 1998.
[20] A. Biswas, et al., Design of fractional-order PIlDm controllers with an improved IEEE World Congress on Computational Intelligence, 1998.
differential evolution, Eng. Appl. Artif. Intell. 22 (2) (2009) 343e350. [31] J. Kennedy, The particle swarm: social adaptation of knowledge, in: IEEE In-
[21] N. Aguila-Camacho, M.A. Duarte-Mermoud, Fractional adaptive control for an ternational Conference on Evolutionary Computation, 1997.
automatic voltage regulator, ISA Trans. 52 (6) (2013) 807e815. € m, B. Wittenmark, Computer-controlled Systems: Theory and Design,
[32] K.J. Åstro
[22] M. Zamani, et al., Design of a fractional order PID controller for an AVR using Courier Dover Publications, 2011.
particle swarm optimization, Control Eng. Pract. 17 (12) (2009) 1380e1387. [33] L. Yun, A. Kiam Heong, G.C.Y. Chong, PID control system analysis and design,
[23] I. Pan, S. Das, Chaotic multi-objective optimization based design of fractional IEEE Control Syst. 26 (1) (2006) 32e41.
order PIlDm controller in AVR system, Int. J. Electr. Power Energy Syst. 43 (1) [34] R.A. Krohling, J.P. Rey, Design of optimal disturbance rejection PID controllers
(2012) 393e407. using genetic algorithms, IEEE Trans. Evol. Comput. 5 (1) (2001) 78e82.
[24] Y. Tang, et al., Optimum design of fractional order PIlDm controller for AVR [35] Y. Del Valle, et al., Particle swarm optimization: basic concepts, variants and ap-
system using chaotic ant swarm, Expert Syst. Appl. 39 (8) (2012) plications in power systems, IEEE Trans. Evol. Comput. 12 (2) (2008) 171e195.
6887e6896. [36] A. Oustaloup, et al., Frequency-band complex noninteger differentiator:
[25] F. Merrikh-Bayat, S.-N. Mirebrahimi, M.-R. Khalili, Discrete-time Fractional- characterization and synthesis, IEEE Trans. Circuits Syst. I Fundam. Theory
order PID Controller: Definition, Tuning, Digital Realization and Experimental Appl. 47 (1) (2000) 25e39.
Results. arXiv preprint arXiv:1405.0144, 2014. [37] J. Wilkie, M. Johnson, R. Katebi, Control Engineering: an Introductory Course,
[26] I. Pan, S. Das, Frequency domain design of fractional order PID controller for Palgrave, 2003.
AVR system using chaotic multi-objective optimization, Int. J. Electr. Power
Energy Syst. 51 (0) (2013) 106e118.