International Journal of Power Electronics and Drive Systems (IJPEDS)

Vol. 14, No. 3, September 2023, pp. 1711~1721

ISSN: 2088-8694, DOI: 10.11591/ijpeds.v14.i3.pp1711-1721  1711

A modified differential evolution algorithm for frequency

management of interconnected hybrid renewable system

Debayani Mishra1, Manoj Kumar Maharana1, Manoj Kumar Kar2, Anurekha Nayak3
1
School of Electrical Engineering, KIIT Deemed to be University, Bhubaneswar, India
2
Electrical and Electronics Engineering Department, Tolani Maritime Institute, Pune, India
3
Department of Electrical Engineering, DRIEMS, Cuttack, India

Article Info ABSTRACT

Article history: In recent decades, the insufficiency in power production has led to the
incorporation of renewable energy sources into microgrid systems. However,
Received Dec 29, 2022 ambiguity in the generation of renewable sources and load variation impacts
Revised Feb 14, 2023 on the system frequency influencing the stable operation of microgrid
Accepted Mar 6, 2023 system. To augment the stable operation of the system an intelligent
controller is required for continuous electric power. The implementation of a
modified differential evolution (MDE)-based cascaded proportional integral
Keywords: derivative fractional filter (PIDFN) controller using integral of time-
weighted absolute error (ITAE) is proposed in this work. In order to
Differential evolution demonstrate the robustness and efficacy of the proposed MDE optimization
Hybrid renewable system technique is compared with differential evolution (DE), teaching learning-
Modified differential evolution based optimization (TLBO), invasive weed optimization (IWO) and particle
Particle swarm optimization swarm optimization (PSO). MDE-based cascaded PIDFN controller is
Teaching learning-based implemented for governing the frequency in two-area interconnected
optimization microgrid system. The system is substantiated over load perturbations,
system uncertainties, communication delay and real time data of solar
irradiance and wind speed and action of unified power flow controller
(UPFC) in MATLAB®/Simulink environment.
This is an open access article under the CC BY-SA license.

Corresponding Author:
Debayani Mishra
School of Electrical Engineering, KIIT Deemed to be University
Bhubaneswar, India
Email: [email protected]

1. INTRODUCTION
Under the notion of sustainable development, the production of renewable energy sources (RESs)
like wind, hydro, and solar has expanded significantly in recent years replacing the traditional thermal power.
The day by day increase in load demand is managed by the power generation from the renewable sources.
The ability of the central grid is strengthened by lowering the peak loads due to evolution of the idea of
decentralization in power generation. In light of this, renewable energy sources are utilized and the concept
of microgrid is implemented. Microgrids are an amalgamation of renewable source, storage units and loads.
The intermittent nature of low inertia renewable sources creates a disparity between generation and demand
which impelled the formation of interconnected microgrid system. Interconnection of microgrid enables the
sharing of surplus power but makes the system more difficult to govern. The frequency change and power
flow in the interconnected tie line gradually deviates from their nominal value when the load changes
dynamically in the interconnected microgrid system. This intensifies severe frequency deviations which
further deteriorates the power quality and hence impacts the microgrid efficiency. In order to balance the

Journal homepage: http://ijpeds.iaescore.com

1712  ISSN: 2088-8694

power transmission across the various power system to maintain a steady grid frequency, load frequency
control is used in micro grids with multiple generating units.
Recent advances in contemporary power networks have necessitated the use of decentralized control
methods rather than centralized control schemes to ameliorate unreliable and inefficient power system control.
The power system frequency was controlled by robust optimal [1], stochastic optimal [2], and secondary loop
frequency control methods [3] in literature. In order to get an enhanced LFC response the controller parameters
need to be tuned rapidly and precisely. In this context, various meta-heuristic optimization techniques like
firefly algorithm [4], ant lion optimizer (ALO) [5], grey wolf optimization [6] and differential evolution [7]
have been proposed in literature. An integrated multi-source, multi-area power system's voltage and frequency
are controlled by a model predictive controller (MPC), whose parameters are tuned via salp swarm algorithm
[8], model predictive controller [9] and using intelligent techniques for energy storage devices [10]. Harris hawk
optimization (HHO) based model predictive controller (MPC) has been implemented for regulating the voltage
and frequency and validated under different conditions in three area hybrid power system as shown in [11] and
[12]. Leader Harris hawk optimization (LHHO) based MPC has been proposed for controlling voltage and
frequency of the system. A hybrid interconnected power system comprising of thermal, wind, diesel,
photovoltaic (PV) and hydrogen units in [13] has been designed. The algorithm was tested in conjunction with
capacitive energy storage units and virtual inertia units. The work in [14] explain an intelligent fuzzy logic
controller was implemented for a multi-area interconnected power system further the controller was compared
with PI and PID subjected to minimization of integral of time-weighted absolute error (ITAE) function.
However, these optimization techniques do not offer exceptional performance in terms of settling time, peak
overshoot or undershoot. Differential evolution (DE) search strategy proposed in [15] was capable to address
optimization problems effectively. In a specific procedure DE performance is dependent on the selection of
values of crossover constant and scaling factor. In this paper modified differential evolution (MDE) technique is
suggested which will overcome the drawback of DE.
From literature it has been studied that the performance of load frequency control (LFC) system
does not depend on the optimization strategy but also rely on the controller design. Different controller such
as adaptive control [16], [17] model predictive control [18] have been proposed to control the frequency in
isolated MG system. These conventional controllers fail to work effectively under various operating
conditions due to aberrant nature of RESs. In contrast, fractional order controllers which improves the
stability of an interconnected MG system have gained a great deal of interest in recent years due to its
versatile structure and has a greater number of tuning parameters. In recent years cascaded controller have
been employed in interconnected power system as it can withstand diverse disturbances effectively. In this
context, a cascaded proportional integral derivative fractional filter (PIDFN) controller is suggested for an
interconnected microgrid system for diminishing the frequency disruption in this paper. The, classical
controller like P, PI and PID works effectively for hybrid MG system. These controllers are sluggish in event
of rapid load disturbance in terms of parameter adjustments. Further sliding mode controllers, fractional order
PI and PID controllers have structural issues and incur more cost. These findings encourage the development
of a cascaded PIDFN controller [19]. The suggested controller has a better performance for abating the
deviations of frequency in the system. Inspired by the research works, this paper suggests a MDE algorithm
for tuning the parameter of the cascaded-PIDFN controller. The suggested optimization techniques surmount
other optimization algorithms in terms of ITAE and convergence.
The contributions and features of the proposed work are as follows: i) A novel modified differential
(MDE) algorithm is used to adjust the parameter of the cascaded PIDFN controller, ii) The supremacy of the
MDE has been established by comparing with other optimization techniques such as differential evolution
(DE), particle swarm optimization (PSO), teaching learning-based optimization (TLBO) and invasive weed
optimization (IWO), iii) The proposed cascaded PIDFN controller performance is studied under various
situations such as variable load disturbance, system uncertainties, communication delay and real time data of
solar irradiance and measurement of wind speed. The actions of unified power flow controller (UPFC) are
also tested on the two-area interconnected microgrid system in MATLAB/Simulink environment, and iv) The
efficacy of optimization technique MDE is verified for different performance parameters.
The article is established as follows: the hybrid renewable energy system (HRES) model is
described in section 2, and the proposed cascaded PIDFN controller is covered in section 3. The performance
of the MDE is discussed in section 4 and simulation results are presented in section 5. Finally, the conclusion
is presented in section 6.

2. MODELLING OF THE PROPOSED SYSTEM

A two-area interconnected microgrid system is designed for analysis of frequency deviation of the
system. The area 1 of the microgrid comprises of a wind turbine generator (WTG), diesel generator (DEG) and

Int J Pow Elec & Dri Syst, Vol. 14, No. 3, September 2023: 1711-1721
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1713

ultracapacitor (ULC) whereas area 2 comprises of a photovoltaic cell (PV), DEG, battery storage as shown in
Figure 1. When the MG system is exposed to load perturbations and the intermittent nature of RES generation
introduces ambiguity in the system which leads to instability. Hence each area required to monitor and control
for proper operation of power system. In this paper a cascaded-PIDFN controller is used, whose parameters are
tuned by MDE algorithm to minimize the frequency deviation and oscillations in each area and tie line.
Additionally, the performance of MDE-PIDFN controller is compared with DE, PSO, TLBO and IWO. The
various components of the proposed MG system are briefly illustrated in the following subsections.

Figure 1. Proposed microgrid system

2.1. Modelling of PV units

The PV system provides fluctuating power due to erratic nature of solar radiation and temperature.
The extracted power from the PV cell is determined using:

𝑃𝑃𝑉 = 𝜂𝑆𝜙(1 − 0.005(𝑇𝑎 + 25)) (1)

where 𝜂= effectiveness of PV array, S = PV surface area, 𝜙 = intensity of solar radiation, Ta = surrounding

temperature.

2.2. Modelling of wind turbine system

The wind turbine's output power is irregular owing to varying wind speed. The WTG is
characterized as a combination of power coefficient ‘Cp’ and other physical factors tip speed ratio ′𝜆′ and
blade pitch angle ‘β’ are the fundamental components of ‘Cp’. The output mechanical power of the WTG is
expressed as:
1
𝑃𝑊𝑇𝐺 = 𝜌𝐴𝐶𝑃 𝑉𝑊 (2)
2

where 𝜌 = is the density of air in kg/m3 and A = is the region swept by the turbine blades in m 3.

2.3. Modelling of diesel engine generator units

Typically, a diesel generator (DG) is installed in places where there is no grid connection, or it is
utilized as an alternate emergency power source when there are grid outages. DG are not only deployed for
backup or emergency units, but they also have auxiliary purposes to offset the intermittent nature of RESs in
remote MG. The diesel engine generator (DEG) is well known in the MG system as an efficient source of
power which acts during load augmentation with great durability and efficiency. The, parameter in the figure

A modified differential evolution algorithm for frequency management of … (Debayani Mishra)

1714  ISSN: 2088-8694

Δf, TGOV, TGE, TDE and R corresponds to the frequency deviation, governor’s time constant, generator time
constant, delay time constant and speed regulation of governor respectively.

2.4. Modelling of battery energy system

The DEG sources inhibits poor response time so they are supplemented by energy storage devices
for improved frequency management. The battery energy system (BES) acts like storage units and frequency
fluctuations are maintained via power exchange with the MG. The transfer function model of BES system is
given by (3).
𝐾𝐵𝐸𝑆
𝐺𝐵𝐸𝑆 (𝑠) = (3)
1+𝑠𝑇𝐵𝐸𝑆

2.5. Modelling of ultracapacitor

It is an electrostatically charged component. Unlike batteries, it can withstand thousands of cycles.
The charging and discharging times of the storage units are significantly faster. It has low resistance and
responds quickly to power fluctuation. The transfer function is described as (4).
1
𝐺𝑈𝐿𝐶 (𝑠) = (4)
1+𝑠𝑇𝑈𝐿𝐶

2.6. Modelling of the microgrid system

MGs consists of distributed energy resources, storage devices, and configurable load blocks to allow
distant grid operations with sufficient control capabilities. In recent years, substantial study has been
undertaken in order to confirm the microgrid's (MG) stability and dependability. Here, in this work, an
isolated MG is analyzed to validate the performance of various frequency regulator controllers. Power, is
supplied by all sources, and a relationship between frequency change and power consumption is established.
The transfer function model is represented as (5).
𝛥𝑓 𝐾𝑃𝑆
= (5)
𝛥𝑃 1+𝑠𝑇𝑃𝑆

3. CONTROLLER DESIGN
To reduce the frequency variation in the MG, a cascaded-PIDFN controller is implemented [19].
The performance of the controller is also compared to that of other controllers under various scenarios. The
suggested controller is a cascaded design of PIDN and PIDFN based on fractional calculus, with
differentiators and integrators. The PIDN controller comprises of P, I, D controller with a derivative filter
coefficient N and the PIDFN controllers is of the form of PI λ Dµ and a derivative filter coefficient N. λ and µ
are the order of integrators and differentiators respectively. With the inclusion of a fractional order PIDN
controller, the integer order PIDN controller may be extended from point to plane, making the PIDN control
method more adaptable and robust [20], [21].The scaling parameters provided by the advanced control
optimization, the gain parameters of a cascaded PIDFN controller becomes more flexible [22], [23] and by
implementing a fuzzy PID filter [24].

4. MDE STRATEGIES
A DE algorithm is proposed which was found to be superior in reference to other evolutionary
algorithm simplicity, ease of implementation, and fewer variables [25]. The operation of DE algorithm
follows four steps: i) initialization, ii) mutation, iii) crossover, and iv) selection. In the first step, it starts with
n numbers of population size and number of decision vectors distributed randomly as, ⃗⃗⃗⃗ 𝑋𝑘𝑙 = (𝑋1𝑙 , 𝑋2𝑙 , . . . , 𝑋𝐷𝑙 )
over D-dimensional search space. In second step, a mutant vector (𝑀 ⃗⃗⃗⃗⃗𝑙 ) is generated by applying the mutation
𝑘
operator any of the following:

DE/rand/1: ⃗⃗⃗⃗⃗
𝑀𝑘𝑙 = ⃗⃗⃗⃗⃗⃗ ⃗⃗⃗⃗⃗⃗𝑙 − ⃗⃗⃗⃗⃗⃗
𝑋𝑘𝑙 1 + 𝑓 (𝑋 𝑘2 𝑋𝑘𝑙 3 ) (6)

DE/best/1: ⃗⃗⃗⃗⃗
𝑀𝑘𝑙 = ⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑙
𝑋𝑏𝑒𝑠𝑡 ⃗⃗⃗⃗⃗⃗𝑙 − ⃗⃗⃗⃗⃗⃗
+ 𝑓 (𝑋 𝑘1 𝑋𝑘𝑙 2 ) (7)

DE/rand/2: ⃗⃗⃗⃗⃗
𝑀𝑘𝑙 = ⃗⃗⃗⃗⃗⃗ ⃗⃗⃗⃗⃗⃗𝑙 − ⃗⃗⃗⃗⃗⃗
𝑋𝑘𝑙 5 + 𝑓 (𝑋 𝑘1
⃗⃗⃗⃗⃗⃗𝑙 − ⃗⃗⃗⃗⃗⃗
𝑋𝑘𝑙 2 ) + 𝑓 (𝑋 𝑘3 𝑋𝑘𝑙 4 ) (8)

DE/best/2: ⃗⃗⃗⃗⃗
𝑀𝑘𝑙 = ⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑙
𝑋𝑏𝑒𝑠𝑡 ⃗⃗⃗⃗⃗⃗𝑙 − ⃗⃗⃗⃗⃗⃗
+ 𝑓 (𝑋 𝑘1
⃗⃗⃗⃗⃗⃗𝑙 − ⃗⃗⃗⃗⃗⃗
𝑋𝑘𝑙 2 ) + 𝑓 (𝑋 𝑘3 𝑋𝑘𝑙 4 ) (9)

Int J Pow Elec & Dri Syst, Vol. 14, No. 3, September 2023: 1711-1721
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1715

DE/current-to-best/1: ⃗⃗⃗⃗⃗
𝑀𝑘𝑙 = ⃗⃗⃗⃗ ⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑋𝑘𝑙 + 𝑓 (𝑋 𝑙 ⃗⃗⃗⃗𝑙 ⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑙 ⃗⃗⃗⃗⃗⃗𝑙
𝑏𝑒𝑠𝑡 − 𝑋𝑘 ) + 𝑓 (𝑋𝑏𝑒𝑠𝑡 − 𝑋𝑘2 ) (10)

⃗⃗⃗⃗𝑙 ) and mutant vector (𝑀

In third step, the decision vector (𝑋 ⃗⃗⃗⃗⃗𝑙 ) undergo crossover operation to form a trial
𝑘 𝑘
𝑙
vector (𝑉𝑖𝑗 ), expressed as (11):
𝑙
𝑙
𝑀𝑘𝑗 𝑖𝑓 𝑟𝑎𝑛𝑑(0,1) < 𝐶𝑟
𝑉𝑘𝑗 =[ 𝑙
𝑓𝑜𝑟 𝑗 = 1,2, … , 𝐷 (11)
𝑋𝑘𝑗 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒.

in last step, the vector for the next generation is selected as per the (12):

𝑉 𝑙 𝑖𝑓 𝑓𝑖𝑡𝑛𝑒𝑠𝑠 𝑜𝑓 𝑉𝑘𝑙 𝑖𝑠 𝑏𝑒𝑡𝑡𝑒𝑟 𝑡ℎ𝑎𝑛 𝑋𝑘𝑙

𝑋𝑘𝑙+1 = [ 𝑘 𝑙 (12)
𝑋𝑘 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

where, ⃗⃗⃗⃗
𝑋𝑘𝑙 is the kth decision vector of lth generation; ⃗⃗⃗⃗⃗
𝑀𝑘𝑙 is the kth mutant vector of lth generation; ⃗⃗⃗⃗
𝑋𝑘𝑙 and
⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑋 𝑙 are the lth generation's randomly chosen decision vector and best solution vector, respectively where f
𝑏𝑒𝑠𝑡
is the scale factor and 𝐶𝑟 is the crossover rate having value between 0 to 1.
The mutation operator like DE/rand mutation operator has more exploration feature than
exploitation which also takes more time to attain optimal solution. Similarly, the DE/best mutation operator
has more exploitation feature than exploration and takes less time to obtain optimal solution. This approach,
however, suffers from early convergence and being stuck in local optima. A novel mutation operator is
presented in the MDE algorithm to prevent the above difficulty and maintain a balance between exploitation
and exploration. Here, six best solutions from the current generation are chosen to find three mutant vectors
in (13) to (15). The final mutant vector (𝑀 ⃗⃗ 𝑘𝑙 ) is then determined by taking the average of these three mutant
vectors. In crossover step, a new trial vector is obtained as (16). Finally, in selection step, the decision vector
for next generation is obtained as expressed in (17).

𝑀1 = ⃗⃗⃗⃗
⃗⃗⃗⃗⃗ ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑋𝑘𝑙 + 𝑓 (𝑋 𝑙 ⃗⃗⃗⃗𝑙 ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑙 ⃗⃗⃗⃗𝑙
𝑏𝑒𝑠𝑡1 − 𝑋𝑘 ) + 𝑓 (𝑋𝑏𝑒𝑠𝑡2 − 𝑋𝑘 ) (13)

𝑀2 = ⃗⃗⃗⃗
⃗⃗⃗⃗⃗ ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑋𝑘𝑙 + 𝑓 (𝑋 𝑙 ⃗⃗⃗⃗𝑙 ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑙 ⃗⃗⃗⃗𝑙
𝑏𝑒𝑠𝑡3 − 𝑋𝑘 ) + 𝑓 (𝑋𝑏𝑒𝑠𝑡4 − 𝑋𝑘 ) (14)

𝑀3 = ⃗⃗⃗⃗
⃗⃗⃗⃗⃗ ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑋𝑘𝑙 + 𝑓 (𝑋 𝑙 ⃗⃗⃗⃗𝑙 ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑙 ⃗⃗⃗⃗𝑙
𝑏𝑒𝑠𝑡5 − 𝑋𝑘 ) + 𝑓 (𝑋𝑏𝑒𝑠𝑡6 − 𝑋𝑘 ) (15)

⃗⃗ 𝑘𝑙 = 𝑀𝑒𝑎𝑛(𝑀
𝑀 ⃗⃗ 1 , 𝑀
⃗⃗ 2 , 𝑀
⃗⃗ 3 ) (16)
𝑙
𝑓 = 2 × (1 − ) (17)
𝐿𝑚𝑎𝑥

𝑙 𝑙 𝑙 𝑙 𝑙 𝑖
Where, 𝑋𝑏𝑒𝑠𝑡1 , 𝑋𝑏𝑒𝑠𝑡2 , 𝑋𝑏𝑒𝑠𝑡3 , 𝑋𝑏𝑒𝑠𝑡4 , 𝑋𝑏𝑒𝑠𝑡5 and 𝑋𝑏𝑒𝑠𝑡6 are the six best decision vectors chosen from lth
generation. 𝑙 and 𝐿𝑚𝑎𝑥 are the current and maximum generation respectively.

5. RESULTS AND DISCUSSION

The proposed MDE algorithm based cascaded-PIDFN controller is tested for frequency management
in a hybrid microgrid system. The robustness of the controller is tested under various condition which includes
arbitrary load disturbances, incorporation of FACTs devices in the hybrid system with real-time solar irradiance
and wind speed data, the effect of communication latency, and parameter variations in the system.
In each case, the cascaded PIDFN controller parameters are tuned by modified differential evolution
(MDE) and the performance is compared with DE, TLBO, PSO and IWO techniques. The comparison is
based on ITAE and dynamic values of ΔF1, ΔF2 and ΔP tie. The population size, swarm size, and number of
search agents in all meta-heuristic algorithms is set at 20 and executed up to 100 iterations. The optimization
algorithms being meta-heuristic in nature so the process has been repeated 10 times independently and ITAE
is measured.

5.1. Effect of variable load performance

The effect of frequency on the variation of load in any area are considered in this case. In this case a
random step load variation in Area 1 as shown in Figure 2 is implemented to test the robustness of the
proposed controller. The, above load profile is applied to the system and the parameters for ITAE are

A modified differential evolution algorithm for frequency management of … (Debayani Mishra)

1716  ISSN: 2088-8694

calculated using different optimization techniques. The different ITAE values obtained from the optimization
techniques are listed in Table 1.
Figures 3(a)-(c) represents the dynamic performance of ΔF1, ΔF2 and ΔPtie respectively under random
load perturbations. From Figure 3(a) it is observed that the transient response of the variation of frequency (ΔF1)
in area 1 has better response using the MDE-PIDFN controller compare to another controller. Similarly in
Figures 3(b) and 3(c) the MDE-PIDFN controller provide better transient response for the variation of
frequency (ΔF2) in area 2 and change in tie line power (ΔP tie) respectively as compared to other controllers
which is shown in the zoom portion of the figure. From Table 2 it is observed that the MDE-PIDFN controller
has a better fitness as compared to other controllers. The transient specification of the system with respect to
peak undershoot (PUS) and settling time (TS) is represented in Table 2. The MDE algorithm offers improved
dynamic responsiveness in terms of settling time as compare to other optimization techniques.

Figure 2. Random load profile

(a) (b)

(c)

Figure 3. Frequency fluctuations a random load perturbation (a) ΔF1, (b) ΔF2, and (c) ΔPtie

Int J Pow Elec & Dri Syst, Vol. 14, No. 3, September 2023: 1711-1721
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1717

Table 1. Comparisons of ITAE value for load fluctuations in Area 1

Controller IWO-PIDFN PSO-PIDFN TLBO-PIDFN DE-PIDFN MDE-PIDFN
ITAE 0.0023 0.0034 0.0022 0.0020 0.00191

Table 2. System specifications under variable load disturbance in Area 1

Controller ΔF1 (Hz) ΔF2 (Hz) ΔPtie (p.u)
PUS (Hz) × 10-4 TS (sec) PUS (Hz) × 10-5 TS (sec) PUS (Hz) × 10-4 TS (sec)
IWO-PIDFN - 4.7 -0.478 Unstable 2.79
TLBO-PIDFN - 5.6 -0.687 Unstable 3.95
PSO-PIDFN -0.48 5.1 -2.25 Unstable -3.15 Unstable
DE-PIDFN -0.1 3.5 -0.756 Unstable -0.12 2.9
MDE-PIDFN - 1.5 -0.65 5 - 0.05

5.2. Effect of the UPFC in the HRES

In this scenario, UPFC is implemented in the system and the microgrid is subjected to real time data
of variation of PV irradiance and wind speed on a particular day which are represented in Figure 4 and
Figure 5. The different ITAE values implementing an UPFC in the HRES is obtained from various
optimization techniques are listed in Table 3.
According to Table 4, the MDE-PIDFN controller has superior fitness compared to other
controllers. The dynamic performance of ΔF1, ΔF2, and ΔPtie under random load perturbations is depicted in
Figures 6(a)-(c). Figure 6(a) demonstrates that the MDE-PIDFN controller provides a superior transient
response to the frequency fluctuation (F1) in region 1 compared to other controllers. Similarly, in
Figures 6(b) and 6(c), the MDE-PIDFN controller provides a better transient response for the frequency
variation (F2) in area 2 and the change in tie line power (ΔP tie) than other controllers, as indicated in the zoom
region of Figure 6. Table 4 displays the transient specifications of the system in terms of peak undershoot
(PUS) and settling time (TS). As comparison to other optimization techniques, the MDE method offers
higher dynamic responsiveness in terms of settling time.

Figure 4. Variation in PV irradiance Figure 5. Variation in wind speed data

Table 3. Different optimization techniques under action of UPFC

Controller IWO-PIDFN PSO-PIDFN TLBO-PIDFN DE-PIDFN MDE-PIDFN
ITAE 0.000007 0.000004 0.0000038 0.0000034 0.0000003

Table 4. System specifications under action of UPFC and random load disturbance in Area 1 and Area 2
Controller ΔF1 (Hz) ΔF2 (Hz) ΔPtie (p.u)
PUS (Hz) × 10-4 TS (sec) PUS (Hz) × 10-6 TS (sec) PUS (Hz) × 10-6 TS (sec)
IWO-PIDFN - 9.65 5.5 - 17.2
TLBO-PIDFN - 7.87 6.78 -0.02 11.6
PSO-PIDFN - 8.23 8.9 - 12.2
DE-PIDFN -1.5 Unstable -2.5 Unstable -0.68 Unstable
MDE-PIDFN - 6.35 4.88 - 9.3

A modified differential evolution algorithm for frequency management of … (Debayani Mishra)

1718  ISSN: 2088-8694

5.3. Robustness of the proposed controller

In order to test the robustness of the proposed controller the HRES is tested under parameters
variations of the power system gain (Kps), power system time constant (Tps) and regulation of the governor
(R) as represented in Table 5 and communication time delay (CTD) of 0.1 secs is implemented at the input of
the controller. The change in operational set point of the generating points may be delayed which impacts on
system stability due to mismatch between generation and demand. CTDs minimizes the system instability so
the influence of CTDs on frequency stability of the HRES is examined here. Furthermore, area 1 is subjected
to a step load change at 30 seconds and area 2 at 20 seconds.

Table 5. Parameter variation

Parameter R Kps Tps
Variation (%) in Area 1 +5 -20 +25
Variation (%) in Area 2 -5 +20 -25

(a) (b)

(c)

Figure 6. Dynamic response under UPFC (a) ΔF1, (b) ΔF2, and (c) ΔPtie

When the HRES system is exposed to variation in parameters and CTD the system ITAE values is
represented in Table 5 for various controllers. From Table 6 it is observed that the ITAE value of the
proposed controller has a better value as compared to other optimization methods. Figures 7(a)-(c) depict the
dynamic response of frequency deviations F1 and F2 in each location, as well as the tie line power variation of
ΔPtie in Areas 1 and 2 respectively. Figure 7(a) demonstrates that the MDE-PIDFN controller provides a
superior transient response to conventional controllers for the frequency fluctuation (F1) in region 1.
Similarly, in Figures 7(b) and 7(c), the MDE-PIDFN controller provides a better transient response for the
frequency variation (F2) in area 2 and the change in tie line power (Ptie) than other controllers, as seen in the
zoomed region. The transient specification of the HRES under this scenario is represented in Table 7. From
Table 7 it is examined that the MDE method provides enhanced dynamic responsiveness in terms of settling
time in comparison to other optimization strategies.

Table 6. ITAE values of different optimization techniques under action of CTD and variation in parameters
Controller IWO-PIDFN PSO-PIDFN TLBO-PIDFN DE-PIDFN MDE-PIDFN
ITAE 0.005945 0.004812 0.004218 0.002984 0.002105

Int J Pow Elec & Dri Syst, Vol. 14, No. 3, September 2023: 1711-1721
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1719

Table 7. System specifications under parameter variation and CTD

Controller ΔF1 (Hz) ΔF2 (Hz) ΔPtie (p.u)
PUS (Hz) × 10-3 TS (sec) PUS (Hz) TS (sec) PUS (Hz) × 10-3 TS (sec)
IWO-PIDFN -0.5 7.8 -0.5 4.97 -0.12 3.85
TLBO-PIDFN -2.12 9.6 -2.12 4.32 -0.62 9.91
PSO-PIDFN -0.8 Unstable -0.8 3.7 - Unstable
DE-PIDFN -0.41 14.98 -0.034 3.8 -0.1 16.3
MDE-PIDFN -1 2.98 - 2.91 - 2.68

(a) (b)

(c)

Figure 7. Dynamic response under CTD and parameter variation (a) ΔF 1, (b) ΔF2 and (c) ΔPtie

6. CONCLUSION
An interconnected microgrid system constituting wind generator, photovoltaic, diesel engine
generator, battery and capacitor is modelled in this paper. The HRES comprising of nature dependent sources
results in frequency perturbations in the system which can be ameliorated by using a controller. In this paper
a MDE technique is suggested to tune the parameters of the cascaded-PIDFN controller of two area MG
system. To assess the effectiveness of the proposed controller the microgrid is subjected to load perturbation,
parametric variations, effect of CTD and UPFC in the system. The dynamic behavior of the MG system is
analyzed and compared with the results of DE, PSO, TLBO and IWO tuned cascaded-PIDFN controller. In
Area 1 the proposed controller improves the settling time by 80.10%, 68.95%, 61.79% as compared to DE-
PIDFN, TLBO-PIDFN and IWO-PIDFN controller. Similarly in Area 2, the settling time is improved by
23.42%, 21.29%, 32.63% and 41.44% in contrast with DE-PIDFN, PSO-PIDFN, TLBO-PIDFN and IWO-
PIDFN controller. The settling time in the tie line is also improved by 83.55%, 72.95% and 30.38% as
compared to DE-PIDFN, TLBO-PIDFN and IWO-PIDFN controller. The proposed controller is accountable
for maintaining frequency fluctuation and making the power system more stable with reduced peak
overshoot, undershoot, and muted oscillations.

A modified differential evolution algorithm for frequency management of … (Debayani Mishra)

1720  ISSN: 2088-8694

REFERENCES
[1] H. Li, X. Wang, and J. Xiao, “Adaptive event-triggered load frequency control for interconnected microgrids by observer-based
sliding mode control,” IEEE Access, vol. 7, pp. 68271–68280, 2019, doi: 10.1109/ACCESS.2019.2915954.
[2] A. Rafinia, J. Moshtagh, and N. Rezaei, “Stochastic optimal robust design of a new multi-stage under-frequency load shedding
system considering renewable energy sources,” Int. J. Electr. Power Energy Syst., vol. 118, 2020, doi:
10.1016/j.ijepes.2019.105735.
[3] H. Gozde, M. Cengiz Taplamacioglu, and I. Kocaarslan, “Comparative performance analysis of Artificial Bee Colony algorithm
in automatic generation control for interconnected reheat thermal power system,” Int. J. Electr. Power Energy Syst., vol. 42, no.
1, pp. 167–178, 2012, doi: 10.1016/j.ijepes.2012.03.039.
[4] N. K. Gupta, M. K. Kar, and A. K. Singh, “Load frequency control of two-area power system by using 2 degree of freedom PID
controller designed with the help of firefly algorithm,” in Control Applications in Modern Power System, vol. 710, 2020, Springer
Singapore, 2021.
[5] M. Raju, L. C. Saikia, and N. Sinha, “Automatic generation control of a multi-area system using ant lion optimizer algorithm
based PID plus second order derivative controller,” Int. J. Electr. Power Energy Syst., vol. 80, pp. 52–63, 2016, doi:
10.1016/j.ijepes.2016.01.037.
[6] Y. Sharma, and L. C. Saikia, “Automatic generation control of a multi-area ST - thermal power system using grey wolf optimizer
algorithm based classical controllers,” Int. J. Electr. Power Energy Syst., vol. 73, pp. 853–862, 2015, doi:
10.1016/j.ijepes.2015.06.005.
[7] M. K. Kar, S. Kumar, A. K. Singh, and S. Panigrahi, “Reactive power management by using a modified differential evolution
algorithm,” Optim. Control Appl. Methods, no. October, pp. 1–20, 2021, doi: 10.1002/oca.2815.
[8] A. Singh, and V. Sharma, “Salp swarm algorithm-based model predictive controller for frequency regulation of solar integrated
power system,” Neural Comput. Appl., vol. 31, no. 12, pp. 8859–8870, 2019, doi: 10.1007/s00521-019-04422-3.
[9] Z. Ji, X. Huang, C. Xu, and H. Sun, “Accelerated model predictive control for electric vehicle integrated microgrid energy
management: A hybrid robust and stochastic approach,” Energies, vol. 9, no. 11, 2016, doi: 10.3390/en9110973.
[10] R. Shankar, K. Chatterjee, and R. Bhushan, “Impact of energy storage system on load frequency control for diverse sources of
interconnected power system in deregulated power environment,” Int. J. Electr. Power Energy Syst., vol. 79, pp. 11–26, 2016,
doi: 10.1016/j.ijepes.2015.12.029.
[11] H. Shukla, S. Nikolovski, M. Raju, A. S. Rana, and P. Kumar, “SMES-GCSC coordination for frequency and voltage regulation
in a multi-area and multi-source power system with penetration of electric vehicles and renewable energy sources,” Energies, vol.
16, no. 1, 2023, doi: 10.3390/en16010251.
[12] M. A. Sobhy, M. Ezzat, H. M. Hasanien, and A. Y. Abdelaziz, “Harris hawks algorithm for automatic generation control of
interconnected power systems,” 2019 21st Int. Middle East Power Syst. Conf. MEPCON 2019 - Proc., 2019, pp. 575–582, doi:
10.1109/MEPCON47431.2019.9007968.
[13] J. Zhao and J. Wang, “Integrated model predictive control of hybrid electric vehicles coupled with aftertreatment systems,” IEEE
Trans. Veh. Technol., vol. 65, no. 3, pp. 1199–1211, 2016, doi: 10.1109/TVT.2015.2405918.
[14] J. Brest, S. Greiner, B. Bošković, M. Mernik, and V. Zumer, “Self-adapting control parameters in differential evolution: A
comparative study on numerical benchmark problems,” IEEE Trans. Evol. Comput., vol. 10, no. 6, pp. 646–657, 2006, doi:
10.1109/TEVC.2006.872133.
[15] A. K. Qin, V. L. Huang and P. N. Suganthan, “Differential evolution algorithm with strategy adaptation for global numerical
optimization,” in IEEE Transactions on Evolutionary Computation, vol. 13, no. 2, pp. 398−417, April 2009, doi:
10.1109/TEVC.2008.927706.
[16] S. Zheng, X. Tang, B. Song, S. Lu, and B. Ye, “Stable adaptive PI control for permanent magnet synchronous motor drive based
on improved JITL technique,” ISA Trans., vol. 52, no. 4, pp. 539–549, 2013, doi: 10.1016/j.isatra.2013.03.002.
[17] H. Bevrani, and P. R. Daneshmand, “Fuzzy logic-based load-frequency control concerning high penetration of wind turbines,”
in IEEE Systems Journal, vol. 6, no. 1, pp. 173−180, March 2012, doi: 10.1109/JSYST.2011.2163028.
[18] J. Yang, Z. Zeng, Y. Tang, J. Yan, H. He, and Y. Wu, “Load frequency control in isolated micro-grids with electrical vehicles
based on multivariable generalized predictive theory,” Energies, vol. 8, no. 3, pp. 2145–2164, 2015, doi: 10.3390/en8032145.
[19] D. Mishra, M. K. Maharana, and A. Nayak, “Frequency amelioration of an interconnected microgrid system,” Open Eng., vol. 12,
no. 1, pp. 349–358, 2022, doi: 10.1515/eng-2022-0038.
[20] R. Rajbongshi and L. C. Saikia, “Combined voltage and frequency control of a multi-area multisource system incorporating dish-
Stirling solar thermal and HVDC link,” IET Renew. Power Gener., vol. 12, no. 3, pp. 323–334, 2018, doi: 10.1049/iet-
rpg.2017.0121.
[21] R. Rajbongshi and L. C. Saikia, “Combined control of voltage and frequency of multi-area multisource system incorporating
solar thermal power plant using LSA optimised classical controllers,” IET Gener. Transm. Distrib., vol. 11, no. 10, pp. 2489–
2498, 2017, doi: 10.1049/iet-gtd.2016.1154.
[22] S. Oladipo, Y. Sun, and Z. Wang, “Maiden application of a modified HBA-optimized cascaded (2DOF + FOPIDN)-PD controller
for load frequency control of diverse source power system,” Math. Probl. Eng., vol. 2022, 2022, doi: 10.1155/2022/4756347.
[23] N. K. Gupta, M. K. Kar, and A. K. Singh, “Design of a 2-DOF-PID controller using an improved sine–cosine algorithm for load
frequency control of a three-area system with nonlinearities,” Prot. Control Mod. Power Syst., vol. 7, no. 1, 2022, doi:
10.1186/s41601-022-00255-w.
[24] Y. Arya, “AGC of two-area electric power systems using optimized fuzzy PID with filter plus double integral controller,” J.
Franklin Inst., vol. 355, no. 11, pp. 4583–4617, 2018, doi: 10.1016/j.jfranklin.2018.05.001.
[25] L. Kumar, M. K. Kar, and S. Kumar, “Statistical analysis based reactive power optimization using improved differential
evolutionary algorithm,” Expert Syst., vol. 40, no. 1, pp. 1–31, 2023, doi: 10.1111/exsy.13091.

Int J Pow Elec & Dri Syst, Vol. 14, No. 3, September 2023: 1711-1721
Int J Pow Elec & Dri Syst ISSN: 2088-8694  1721

BIOGRAPHIES OF AUTHORS

Debayani Mishra received her B.Tech and M.Tech degree from Biju Pattanaik
University of Technology, Rourkela, Odisha. She is presently working as an Asst. Professor in
Electrical Engineering Department at Ajay Binay Institute of Technology, Cuttack, Odisha.
She is currently pursuing her PhD in KIIT Deemed to be University, Bhubaneswar, Odisha.
She has 14 years of teaching experience her research area includes renewable energy, power
system and energy management system. She can be contacted at email:
[email protected].

Manoj Kumar Maharana is currently Associate Professor in school of Electrical

Engineering, KIIT Deemed to be University, Bhubaneswar, Odisha, India. Before this he was
working as Associate Professor at VIT University Vellore and software developer at GECE
Hyderabad. He received his Ph.D in Electrical Engineering from Indian Institute of
Technology Madras, India in the year 2010. Prior to this M.Tech in power system from NIT-
Warangal(RECW) in 2001 and bachelor degree from The Institution of Engineers, (India),
Kolkata, in the year-1997. During his research he received the Corps of Electrical and
Mechanical Engineers' Prize 2010-2011, for the paper "Direct Acyclic Graph Based Corrective
Control Strategies for Alleviation of Line Overloads during Contingencies He has 17+ years of
teaching and research experience, and guided many M.Tech and PhD Students. He published
several papers in international and national journal and attained several national and
international conferences. His areas of research interest are computer modeling of power
systems, Energy management system, Energy storage and monitoring and Smart grid. He can
be contacted at email: [email protected].

Manoj Kumar Kar is currently working as a Senior Assistant Professor in

Electrical and Electronics Engineering department, Tolani Maritime Institute, Pune, India. He
received his Ph.D in Electrical Engineering from National Institute of Technology,
Jamshedpur, India in the year 2022; M Tech in Power Electronics & Control Drives from Veer
Surendra Sai University of Technology (VSSUT), Burla, India in the year 2014; and B Tech
from Biju Patnaik University of Technology (BPUT), Odisha, India in the year 2010. He has
10+ years of teaching and research experience. He published several research papers in
international Journal and Conferences. His research area includes power system stability,
optimal reactive power dispatch, multilevel inverter, and application of FACTS controllers. He
can be contacted at email: [email protected].

Anurekha Nayak is currently Associate Professor in Department of Electrical

Engineering, DRIEMS, Cuttack, Odisha, India. She received her Ph.D in Electrical
Engineering from KIIT Deemed to be University, Bhubaneswar, Odisha, India in the year 2022.
Prior to this M.Tech in Electrical Power System from Biju Pattanaik University & Technology
in 2011 and bachelor degree from Utkal University, Odisha, India in the year 2002. She has
16+ years of teaching experience and guided many M.Tech students. She published several
papers in international and national journal and attained several national and international
conferences. Her areas of research interest are load frequency control, hybrid power system
and renewable energy. She can be contacted at email: [email protected].

A modified differential evolution algorithm for frequency management of … (Debayani Mishra)