Electric Power Systems Research 220 (2023) 109381

Contents lists available at ScienceDirect

Electric Power Systems Research

journal homepage: www.elsevier.com/locate/epsr

Fractional order based on a flower pollination algorithm PID controller and

virtual inertia control for microgrid frequency stabilization
Komsan Hongesombut, Ruangyos Keteruksa ∗
Department of Electrical Engineering, Faculty of Engineering, Kasetsart University, 50 Ngamwongwan Rd., Ladyao, Chatuchak, Bangkok 10900, Thailand

ARTICLE INFO ABSTRACT

Keywords: In this study, a fractional order PID (FOPID) controller is used to reduce the relative problems and stabilize the
Renewable energy sources power system. However, an intrinsic disadvantage of using FOPID controllers is that generating or adjusting
Stability of the power system the correct settings is time-consuming and is ineffective in producing the desired results. The issue can be
Fractional order PID controller
resolved by determining the best set of parameters for the FOPID controller using a pollination method. As
Flower pollination algorithm
part of a thorough strategy to attain dynamic stability, the FOPID controller changes the influence of the
Virtual inertia control
Microgrid frequency stabilization
power system on the transient and stable state stabilities, and balances the necessary load power in response
to the productivity of all the generators. The modeling approach was divided into two primary phases, and
the dynamic behavior of a microgrid system was simulated using a virtual inertia control model. An s-plane
stability analysis-based non-metric theory, or a PID controller model with an s-domain and fractional order,
were used for the study. The simulation and analytical models were performed using the FOMCON Toolbox.
This study focuses on the impact of adding renewable energy sources on microgrid systems. The fundamental
conclusion of this study is that the recommended virtual inertia control paradigm improves the dependability,
economic value, and sustainability of the energy production and distribution.

1. Introduction also effects renewable energy sources, which are often found to have
a significantly small latent inertia. Based on these issues, this study
The desired primary power systems must be grid-flexible, supplying focuses on two main parts to obtain the right combination. First,
loads to satisfy power needs under dependable and stable conditions. we used virtual inertia [1–8] to replace the inertia required for a
Electricity usage is rapidly and steadily increasing over time. On the power system, which is designed according to the frequency and power
other hand, the use of coal and fuel oil, which are the primary energy control standards. Second, a fractional-order PID controller [9–15]
sources, is decreasing and is insufficient for the future. Renewable was selected to control the frequency deviation. Slowdowns should be
energy is an exciting and popular alternative to fossil fuels. Most of alleviated and supported to enter a stable state, for which the control
this renewable energy comes from renewable energy sources such as signal should not be excessively large, preventing actuator saturation
wind and solar energy, which are clean energy sources that exist in and reducing costs associated with larger actuators. Remote control
nature; however, electricity is obtained from wind power (WP) and signals and a high latency require a ramp-rate response for faster loads.
solar power (PV) generators. The inconsistent variations in the amount The relevant studies are divided into two main topics. The first topic
of sun and wind naturally available prevent homogeneity. This harms focuses on the study [1–8,16–27] of controlling the virtual inertia by
the stability of a power system owing to oscillations in the frequency
presenting the synthesis aspect, most of which is aimed at the robust
and power. Another factor in these power sources is the growing role
stability and control issues related to the system inertia dynamics
of renewable energy. Therefore, certain main larger generators were
caused by the inclusion of DG/RESs in a power system. This is also
scaled back and decommissioned, decreasing the system inertia and
an application of various forms of inertia. Each article provides a
easily destabilizing the system when an electrical fault condition or
comprehensive and, detailed description and analysis of several tradi-
interference occurs. Inertia is an important component for the stability
tional and modern control methods to improve the stability and control
of a power system because it helps eliminate frequency deviations by
efficiency of the power system. The second topic analyzes the controls
maintaining a slowdown to enter a steady state and stabilizing the
frequencies from oscillations within a short period. However, inertia designed by using the fractional order PID (FOPID) based on the flower

∗ Corresponding author.
E-mail addresses: [email protected] (K. Hongesombut), [email protected] (R. Keteruksa).
URLs: http://www.ku.ac.th (K. Hongesombut), http://www.ku.ac.th (R. Keteruksa).

https://doi.org/10.1016/j.epsr.2023.109381
Received 15 November 2022; Received in revised form 19 February 2023; Accepted 29 March 2023
Available online 11 April 2023
0378-7796/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
K. Hongesombut and R. Keteruksa Electric Power Systems Research 220 (2023) 109381

of the study from the four scenarios are presented in the following
Nomenclature sections: step change and random load. The total power generated
ACE Area Control Error by the random load of the wind turbines changes the scenario and
AGC Automatic Generation Control situational sensitivity analysis. The results obtained using the FOPID
DG Distributed Generators controller demonstrated that the frequency deviation and wired power
fluctuations were alleviated more effectively than the PID controller. In
DSA Dragonfly Search Algorithm
2019, Babaei and Safari [11] used SCA to adjust the FOPID regulator
ESS Energy Storage System
coefficient, with certain LFC efficacy assessments indicated by IAE. The
EPS Electric Power System simulations were performed using single-and dual-area LFC systems via
EVs Electric Vehicles EV collection. With various delays, the proposed controller presented
FOPI Fractional order Proportional Integral a smaller frequency variation than others. These results can be used as
FOPD Fractional order Proportional Derivative a guide to evaluate the efficacy. A regulator was proposed to reduce
FOPID Fractional-order Proportional Integral the frequency oscillations in the LFC caused by unstable loads. In
Derivative Control 2020, Emre elik [10] introduced FOPI and FOPD controllers, where
FPA Flower Pollination Algorithm the systems and differentiators can be fractional. The gain parame-
GRC Generator Rate Constraint ters and control fractal order were simultaneously adjusted using the
HPs Heat Pumps proposed DSA to reduce the ITAE deviations. This study extended
to single-source/multi-area EPSs. Several comparisons, such as DSA-
IAE Integral Absolute Error
optimized FOPID controls and FOPI and FOPD controller results, are
IEEE Institute of Electrical and Electronics Engi-
used to mitigate the frequencies, undertime, overtime, and tie-line
neers
power oscillations. In 2021, Shen et al. [32] proposed a control strategy
ITAE Integral Time Absolute Error
for aluminum electrolyte loads to participate in the FOPID frequency
ITSE Integral of Time multiplied Squared Error modulation. It was also designed and compared with the conventional
LFC Load Frequency Control PID controller and algorithms. The PSO optimizes the parameters of
PEI Power Electronics Interface the FOPID controller. A two-region LFC model was created to validate
PID Proportional Integral Derivative controllers an integrated control strategy to support high-frequency aluminum
PSO Particle Swarm Optimization electrolyte loads. The output of the differential frequency control sat-
PSS Power System Stabilizer isfies these requirements. The system frequency recovery time was
PV Solar Power demonstrated to be shorter when an aluminum electrolyte load was
RESs Renewable Energy Sources applied to the LFC. The secondary frequency control of the system is
also better.
SCA Sine Cosine Algorithm
The results of a study form five years ago revealed that FOPID
TID Tilted Integral Derivative controllers
controllers offer superior controllability over traditional controllers,
WP Wind Power
and for FOPID controller applications in power systems. It was found
that the FOPID controller was able to adjust the control signal, and is
also flexible. Consequently, the oscillation frequency in the LFC system
is reduced. In the case of an unstable load, the signal settling, minimum,
pollination algorithm (FPA) proposed in several previous studies [9–
and overload times are alleviated. It also requires a shorter time to
15,28–30]. These controls are parameter optimization problems based
sufficiently recover the stability of the frequency in the event of an
on constraints that can be solved using powerful metaheuristics. A
interference in various forms.
survey of related research found that the pollination step algorithm
The contributions of the aforementioned studies can be summarized
[12] is one of the most powerful sensing metrics with a simple al-
as follows:
gorithm that is naturally inspired. The FPA has been proven to have
broad convergence properties (global convergence) [12] and has been 1. A fractional-order PID controller based on flower pollination was
applied to solve a wide range of engineering problems, especially in proposed to solve the frequency problem caused by the use of
control system engineering. The main reason for choosing the FOPID renewable energy sources in microgrid systems.
controller [15] for the system analyzed in this study was because the 2. A virtual inertia simulation was created to replace the dynamic
FOPID controller has five parameters for tuning, presenting a compre- behavior of microgrid systems in the event of interference from
hensive performance beyond the traditional controllers. Therefor, the various sources.
hypothesis proposed in this study can accurately describe the dynamic
behavior of a system [30]. Research findings related to the meta-study 2. System overview and modeling
application of PID controllers used to design the FOPID controllers
for the LFCs in microgrid frequency stabilization are presented. The 2.1. Microgrid system
modern design details of the FOPID found in previous literature are
summarized in Table 1. The microgrid with RESs penetration employed in this study is
In 2017, Morsalia et al. [31] used FOPID and TID controls for shown in Fig. 1. This system consists of 20-MW thermal power plant
the design of an LFC. In this study, we chose a particle cluster con- stations, 8.5-MW solar power plants, 7.5-MW wind power plants, 5-MW
trol parameter search using an optimization algorithm. The results residential loads, 10-MW industrial loads, and 4.5-MW energy storage
demonstrated that FOPID-based LFCs provide a better performance system (ESS); the system base is 15-MW [33]. The primary electrical
than that of the TID and IOPID LFCs in region-frequency damping and system being studied and the energy level in each unit are decentralized
tie-line power oscillations because the transient behavior of the system for microgrid systems. Communication connections (solid lines) link
was improved in terms of the amplitude of oscillation, vibration, and the electrical equipment and present the data and status in different
settling time. In 2018, Rahmat Khezri et al. [28] presented FOPID places. The units for the primary and secondary control loops and
regulators for the power regulation in HPs and the discharge states inertia were controlled by the network for frequency regulation (dashed
in EVs and AGCs to demonstrate the efficiency of the proposed SCA- line). Finally, the power network (dotted line) supplies energy to the
based FOPID controllers in hybrid dual-area power systems. The results microgrid system unit.

2
K. Hongesombut and R. Keteruksa Electric Power Systems Research 220 (2023) 109381

Table 1
State-of-the-art design of FOPID controllers found in previous literature.
Authors Implemented Results
[31] Javad Morsalia et al, 2017 Fractional Order Proportional Integral Derivative (FOPID) FOPID-based LFCs achieved the highest dynamic performance
and Tilted Integral Derivative (TID) controllers in terms of the largest lowest latency ratio, and smallest
ITSE; the smallest maximum value and precipitation time of
the area frequency and oscillation of the wired energy.
[28] Rahmat Khezri et al, 2018 Fractional Order Proportional Integral Differential (FOPID) FOPID controller reduced the frequency deviation and wired
controllers power fluctuations were alleviated more effectively compared
to PID controllers.
[11] F. Babaei and A. Safari, 2019 The Sine Cosine Algorithm (SCA) to adjust the FOPID The results can be used as a guide for evaluating the
controller coefficient effectiveness of the proposed controller for reducing the
frequency oscillations in the LFC system caused by unstable
loads. This reduces the undershot and overtime and sets the
time of the signal error.
[10] Emre elik, 2020 Fractional Order Proportional Integral–Fractional Order A DSA optimized FOPI-FOPD cascade controller for
Proportional Derivative (FOPI-FOPD) cascade controller alleviating the settling time, undershoot, and overshoot of
the frequency and tie-line power oscillations was proposed.
[32] Xin Shen et al, 2021 Fractional Order PID (FOPID) controllers are designed and The different demand responses and random disturbances
compared with traditional PID has proposed a control demonstrate that when an aluminum electrolyte load is
strategy for electrolytic aluminum loads applied in the LFC, the system frequency recovery time is
shorter and the dynamic performance of the secondary
frequency regulation of the system also improves.

Fig. 1. The microgrid system employed in this study.

Fig. 2 depicts the dynamic model of the microgrid under investiga- Table 2
tion, which presents the block diagrams of a typical frequency control The microgrid dynamic and control parameters.

study including the GRC for the governor unit and Parameters Value
( the rate
) limitation of
the turbine-valve/gate closing or opening speed 𝑉𝑈 , 𝑉𝐿 for the turbine Microgrid control system
unit. The GRC was estimated to be 12% p.u. MW/min in this study. The System inertia, H (p.u. MWs) 0.083
System damping coefficient, D (p.u. MW/Hz) 0.015
dead-band limit was set as 0.06% (±0.035 Hz), and 2 s was designated
as the time delay. The parameters of the studied microgrid system are Inertia control unit
listed in Table 2 [14]. Virtual inertia value, 𝐽𝑉 𝐼 (s) 1.6
Virtual damping value, 𝐷𝑉 𝐼 (s) 1.2
Considering the inertia, primary, and secondary control, as well as Time constant of inverter-based ESS, 𝑇𝐸𝑆𝑆 (s) 10
the dynamic effects of the generations and loads shown in Fig. 2, the Maximum capacity of ESS, 𝑃𝐸𝑆𝑆𝑚𝑎𝑥 (p.u. MW) 0.3
incremental change (𝛥𝑃𝑇 𝑂𝑇 𝐴𝐿 ) of the total power outputs is obtained Minimum capacity of ESS, 𝑃𝐸𝑆𝑆𝑚𝑖𝑛 (p.u. MW) −0.3
as follows: Primary control unit
Droop constant, R (Hz/p.u. MW) 2.4
𝛥𝑃𝑇 𝑂𝑇 𝐴𝐿 = 𝛥𝑃𝑚 + 𝛥𝑃𝑊 𝑇 + 𝛥𝑃𝑆𝑉 − 𝛥𝑃𝑖𝑛𝑒𝑟𝑡𝑖𝑎 − 𝛥𝑃𝐿 (1)
Time constant of the governor, 𝑇𝑔 (s) 0.1
The frequency and power variation of the RESs system can be Time constant of a turbine, 𝑇𝑡 (s) 0.4
calculated as follows, considering the influence of the controllers and Secondary control unit
inertia controls: Frequency bias factor, 𝛽 (p.u. MW/Hz) 1.0
1 ( ) Integral control variable gain, 𝐾𝑖 (s) 0.05
𝛥𝐹 = 𝛥𝑃𝑚 + 𝛥𝑃𝑊 𝑇 + 𝛥𝑃𝑆𝑉 − 𝛥𝑃𝑖𝑛𝑒𝑟𝑡𝑖𝑎 − 𝛥𝑃𝐿 (2) Renewable
2𝐻𝑠 + 𝐷
Time constant of the wind turbine, 𝑇𝑊 𝑇 (s) 1.5
when,
Time constant of the solar system, 𝑇𝑃 𝑉 (s) 1.85
1 ( )
𝛥𝑃𝑚 = 𝛥𝑃𝐺 (3)
1 + 𝑠𝑇𝑡
( )
1 1
𝛥𝑃𝐺 = 𝛥𝑃𝐴𝐶𝐸 − 𝛥𝐹 (4)
1 + 𝑠𝑇𝑔 𝑅

3
K. Hongesombut and R. Keteruksa Electric Power Systems Research 220 (2023) 109381

Fig. 2. The dynamic structure of the microgrid used in the frequency response analysis model of this study.

−𝐾𝑖
𝛥𝑃𝐴𝐶𝐸 = (𝛽𝛥𝐹 ) (5)
𝑠

𝛥𝑃𝑅𝐸𝑆 = 𝛥𝑃𝑊 𝑇 + 𝛥𝑃𝑃 𝑉 (6)

1 ( )
𝛥𝑃𝑊 𝑇 = 𝛥𝑃𝑤𝑖𝑛𝑑 (7)
1 + 𝑠𝑇𝑊 𝑇

1 ( )
𝛥𝑃𝑃 𝑉 = 𝛥𝑃𝑠𝑜𝑙𝑎𝑟 (8)
1 + 𝑠𝑇𝑃 𝑉

𝛥𝑃𝑃 𝑉 = 𝛥𝑃𝑆𝐿 + 𝛥𝑃𝐵𝐿 (9)

where,
𝛥𝑃𝑚 is the change in the power produced by the thermal power
plant,
Fig. 3. The factor that converts the rotor speed to grid frequency.
𝛥𝑃𝑖𝑛𝑒𝑟𝑡𝑖𝑎 is the virtual inertia power change of the ESS,
𝛥𝑃𝐿 is the overall load change of the system,
𝐻 is the corresponding system inertia or microgrid inertia,
3.1. Optimal virtual inertia control concept
𝐷 is the comparable microgrid damping coefficient,
𝛥𝑃𝐺 is the governor unit-based power variation of the thermal The active power of the ESS is assumed to be inversely proportional
power plant, to the derivative of the grid frequency through the power electronics
𝛥𝑃𝐴𝐶𝐸 is the secondary control signal change, interface (PEI). In this case, we simulated the virtual inertia in the
𝛥𝑃𝑅𝐸𝑆 is the overall change in the power provided by the renew- power system, which improved the inertial response to variations in the
able energy sources, power demand by increasing the conversion factor between the rotor
𝛥𝑃𝑃 𝑉 is the fluctuation of the output electricity of the solar power speed and grid frequency. The number of poles of the machine, virtual
plant, inertia, and nominal grid frequency contributed to the system (see the
dynamic structure in Fig. 3). Virtual inertia is initiated by receiving
𝐾𝑖 is the secondary frequency controller,
an impulse to behave dynamically when an electric current crosses
𝛥𝑃𝑤𝑖𝑛𝑑 is the first power change due to wind speed,
an electromagnetic pole. It is inducted at the north and south poles,
𝛥𝑃𝑠𝑜𝑙𝑎𝑟 is the first shift in the solar radiation power, causing a repulsion for the electromagnetic pole to cross the magnetic
𝛥𝑃𝑆𝐿 is the change in load power of the residential area, field based on the size of the repulsion force and is introduced in reverse
𝛥𝑃𝐵𝐿 is the change in load power of the industrial area. to the size of the electric power. This behavior allowed us to consider
the position of the power balance. Considering that the fault current
3. Methodology causes the power-balance position to move out of its original position,
we can increase or decrease the control size.

The implementation can be divided into the following four main 3.2. The virtual inertia control structure for microgrid frequency stabiliza-
areas: (1) optimal virtual inertia control concept, (2) virtual inertia con- tion
trol structure based on the frequency response model, (3) virtual inertia
control methods, and (4) Matignon’s theorem-based stability analysis, The study of frequency control is based on the IEEE standard dual-
all of which are supplemented by other algorithms as indicated below. region LFC model; Fig. 2 displays the dynamic model of the examined

4
K. Hongesombut and R. Keteruksa Electric Power Systems Research 220 (2023) 109381

𝑒𝑠𝑠 is a steady-state error, and 𝑒𝑠𝑠−𝑚𝑎𝑥 is the maximum allowance of 𝑒𝑠𝑠 .

The design of the FPA algorithm for the FOPID is broken down into the
following steps:
𝑴𝒊𝒏𝒊𝒎𝒊𝒛𝒆 𝛥𝐹 (𝐾𝑝 , 𝐾𝑖 , 𝐾𝑑 , 𝜆, 𝜇) ⎫
𝑆𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 𝐾𝑝−𝑚𝑖𝑛 ≤ 𝐾𝑝 ≤ 𝐾𝑝−𝑚𝑎𝑥 , ⎪ ⎪
𝐾𝑖−𝑚𝑖𝑛 ≤ 𝐾𝑖 ≤ 𝐾𝑖−𝑚𝑎𝑥 , ⎪
𝐾𝑑−𝑚𝑖𝑛 ≤ 𝐾𝑑 ≤ 𝐾𝑑−𝑚𝑎𝑥 ,⎪
⎪
𝜆𝑚𝑖𝑛 ≤ 𝜆 ≤ 𝜆𝑚𝑎𝑥 , ⎬ (10)
𝜇𝑚𝑖𝑛 ≤ 𝜇 ≤ 𝜇𝑚𝑎𝑥 , ⎪
⎪
𝑃 𝑂𝑟𝑒𝑔 ≤ 𝑃 𝑂𝑟𝑒𝑔−𝑚𝑎𝑥 , ⎪
𝑡𝑟𝑒𝑔 ≤ 𝑡𝑟𝑒𝑔−𝑚𝑎𝑥 , ⎪
⎪
Fig. 4. The FPA-based FOPID controller design for virtual inertia control.
𝑒𝑠𝑠 ≤ 𝑒𝑠𝑠−𝑚𝑎𝑥 , ⎭

Flower Pollination Algorithm (Pseudo code)

microgrid with RESs penetration. This graphic presents a schematic Objective function min or max 𝛥𝐹 (𝑥), 𝑥 = [𝐾𝑝 , 𝐾𝑖 , 𝐾𝑑 , 𝜆, 𝜇] as equation
line diagram of the nonlinear model of a standard microgrid control (10)
area with a thermal power plant as the traditional producing unit and Initialize a population of n flowers [pollen gametes] with random
wind/solar power plants as the renewable generating unit. Incidentally, solutions
the advantages of the referenced structures can be applied in current Find the best initial solution 𝑔 ∗ = 𝑥 in the initial population
practical applications. However, adding renewable generation units to a Define a proximity probability 𝜌 ∈ [0, 1]
system would change the frequency, causing it to oscillate violently and while (t <MaxGeneration)
resulting in difficulty to restabilize the system. Therefore, this problem for i=1: n (all n flowers in the population)
must focus on an effective and reliable system control method when
if rand <p,
applied to address the future critical issue of remaining robust.
Generate a step vector L from a Lévy flight from equation
(12)
3.3. The FOPID base on FPA and virtual inertia control methods
Generate a new solution 𝑥∗ , 𝑥∗ = [𝐾𝑝 , 𝐾𝑖 , 𝐾𝑑 , 𝜆, 𝜇] from
equation (11)
The best current and outdated methods were verified to provide
else
the best performance. This study proposes a fractional-order PID con-
Draw 𝜀 from a uniform distribution via equation (14)
troller based on the FPA presented in a previous study [12]. It is
Do local pollination via equation (13)
used to change the virtual inertia gain of the control, which produces
end if
a virtual inertial response to absorb the oscillating frequencies, and
Evaluate new solutions 𝑥∗ , by the objective function 𝛥𝐹 (𝑥 ∗
rapidly achieves a steady state for the system on the microgrid. Fig. 4
), in equation (13)
depicts the suggested self-adaptive virtual inertial control. During the
If new solutions are better, update them in the population
low penetration frequency deviation of the RESs, the significantly large
end for
value may lead to a longer stabilization time if the microgrid is used
Find the current best solution 𝑔 ∗= 𝑥 ∗
with the virtual inertia constant. The frequency of the system leads to
end while
a slower vibration.
Output the best solution found
Despite these exceptional circumstances, the system requires a mod-
est virtual inertia constant to cope with the frequency of the system,
( )
resulting in a rapid oscillation. In contrast, a virtual inertia constant is 𝑥∗ = 𝑥 + 𝐿 𝑥 − 𝑔 ∗ (11)
required during the frequency variation with substantial RESs penetra-
tion. As a result, it is critical to limit the large amplitude fluctuations 𝜆𝛤 (𝜆) sin(𝜋𝜆∕2) 1
𝐿≈ , (𝑠 ≫ 𝑠0 > 0) (12)
generated by the penetration, with high RESs correlating to a low 𝜋 𝑠1+𝜆
system inertia, to prevent the instability and collapse of the system. ( )
Consequently, assuming that the virtual inertia control mechanism is 𝑥∗ = 𝑥 + 𝜀 𝑥𝑗 − 𝑥𝑘 (13)
properly implemented, changes in the interference/penetration RESs {
1
can be tracked by adjusting the virtual inertia constant, achieving 𝑏−𝑎
, 𝑎≤𝜌≤𝑏
𝜀 (𝜌) = (14)
an optimal performance and microgrid stability, and calculating the 0, 𝜌 < 𝑎 or 𝜌 > 𝑏
optimal value of the virtual inertia constant will be automatically
Suggestions and precautions to use the proposed method are as
conducted. The PID design method for the virtual inertial control of
follows:
the recommended fractional order based on the FPA can be used to
create a block diagram for a single-area power system, as shown in 1. The problem of the poor overlap of signals between the PSS and
Fig. 4. Assuming a zero-input (r = 0), the FOPID parameters are allotted FOPID controller will be modified with the FPA.
values that are in line with their search spaces and inequality-restricted 2. The traditional proportional integral derivative controllers, PID
locations, as indicated in Eq. (10) 𝛥𝐹 is chosen as the objective function controller application, cannot obtain a poor dynamical perfor-
to be minimized. Here, 𝐾𝑝−𝑚𝑖𝑛 is the lower boundary and 𝐾𝑝−𝑚𝑎𝑥 is mance to capture the design objectives.
the upper boundary of the search spaces of 𝐾𝑝 , 𝐾𝑖−𝑚𝑖𝑛 is the lower
boundary and 𝐾𝑖−𝑚𝑎𝑥 is the upper boundary of the search spaces of 𝐾𝑖 , 3.4. Matignon’s theorem-based stability analysis
𝐾𝑑−𝑚𝑖𝑛 is the lower boundary and 𝐾𝑑−𝑚𝑎𝑥 is the upper boundary of the
search spaces of 𝐾𝑑 , 𝜆𝑚𝑖𝑛 is the lower boundary and 𝜆𝑚𝑎𝑥 is the upper Incidentally, analyzing the stability of the system is the most critical
boundary of the search spaces of 𝜆, 𝜇𝑚𝑖𝑛 is the lower boundary and 𝜇𝑚𝑎𝑥 aspect; that is, the plant to be used must prove that there are no poles or
is the upper boundary of the search spaces of 𝜇, 𝑃 𝑂𝑟𝑒𝑔 is the percent zeros in the region or region of destabilization. In the present microgrid
overshoot of regulation, 𝑃 𝑂𝑟𝑒𝑔−𝑚𝑎𝑥 is a maximum allowance of 𝑃 𝑂𝑟𝑒𝑔 , with RESs penetration simulations and analyses were performed using
𝑡𝑟𝑒𝑔 , is the recovering time, 𝑡𝑟𝑒𝑔−𝑚𝑎𝑥 , is a maximum allowance of 𝑡𝑟𝑒𝑔 , the MATLAB/Simulink program with the FOMCON Toolbox [29]. The

5
K. Hongesombut and R. Keteruksa Electric Power Systems Research 220 (2023) 109381

Fig. 5. The relationship between the Riemann surface and the 𝜔 plane.

fractal stability of the studied system must be checked by considering Table 3

The FOPID controller parameters with optimal solutions.
the characteristic equation of the system when 𝛼𝑖 ∈ R+ is the multi
valued function of the complex variable s-domain of the Riemann Controllers type Load regulation response

surfaces. Matignon’s dependability describes how the fractional-order Kp Ki Kd L M

system remains stable, even when the roots are in the right half of the FOPID 13.465294 38.029416 2.000000 0.036543 0.449970
s-plane, which is demonstrated by Matignon’s stability theorem [34]. Fbest = 9.587412.
The fractional-order transfer function G(s) =Y(s)/R(s) is stable in the Time consumed = 261.8063 s.
s-plane if and only if the following requirements are met: Number of objective valuations = 12 525.

𝜋
|𝑎𝑟𝑔(𝜎)| > 𝑞 , ∀𝜎𝑖 ∈ C, 𝑖𝑡ℎ 𝑟𝑜𝑜𝑡 𝑜𝑓 𝑅(𝜎) = 0, (𝑤ℎ𝑒𝑟𝑒 𝜎 = 𝑠𝑞 ) (15)
2 Table 4
Inertia and load damping parameter.
If the closed-loop characteristic equation has s = 0 as the sole
Percentage System inertia H (p.u.) System load damping D (p.u.)
single root, the system remains unstable; thus, Fig. 5(b) depicts the
stability region of a fractional-order system as defined by Matignon’s 100 0.0830 0.0160
70 0.0581 0.0112
stability theorem [34]. A stability analysis undergoes the following
40 0.0332 0.0064
steps. Consider using the generic fractional order transfer function
characteristic equation given below to assess the stability of a system:
∑
𝑖
𝑣𝑖
𝛼0 𝑠𝛽0 + 𝛼0 𝑠𝛽1 + ⋯ + 𝛼0 𝑠𝛽𝑛 = 𝛼𝑖 𝑠𝛽𝑖 = 0 (𝑤ℎ𝑒𝑟𝑒 𝛽𝑖 = ) (16) fractional-order PID controller provides an excellent variable frequency
𝑛
𝑣 response because it has better parameters covering the problematic
Allow the fractional terms of the characteristic equation to be area than that of the PID controller. In addition, virtual inertia helps
shifted from the s-plane to the 𝛼−plane to make the system of equations to support and stabilize the system until a desired stability of the
in a linear integer order: power system is achieved. The thick solid blue line in the figures below
indicates the base. Additional experimental results are, compared with
∑
𝑛
𝑣𝑖 ∑
𝑖
𝛼𝑖 𝑠 𝑣 = 𝛼𝑖 𝜎 𝑣𝑖 = 0 (17) other random constant units under the circumstances established to test
𝑖 𝑛 the effectiveness of the proposed method.
𝑘 Case 1. Effect of the system inertia and load damping control unit.
In this case, 𝑚 is the least common multiple of 𝑣 and 𝜎 = 𝑠 𝑚 . For In the frequency response test for the microgrid power system, when
any value of 𝛼𝑖 , the roots of (17) are ||𝛷𝜎 || = |𝑎𝑟𝑔(𝜎)| . Thus, the stability the inertia (H) and damping (D) parameters from Eq. (2) change, the
criteria for fractional-order systems are as follows: system is tested under degradation at 100%, 70%, and 40% of the H
𝜋 𝜋 and D changes, as shown in Table 4, to examine how the H and D
1. If 2𝑚
< |𝑎𝑟𝑔(𝜎)| < 𝑚
, the system is stable.
𝜋 systems behave given the proposed model to obtain the best response
2. The system is oscillatory stable if |𝑎𝑟𝑔(𝜎)| = 2𝑚 possible. The experimental results are shown in Fig. 6.
Otherwise, the system would be unstable. The aforementioned Case 2. Effect of the virtual inertia droop control.
fractional-order method of stability can be projected into a limited The frequency response test of the models of the thermal power
number of discs in this study. These were applied to 100 sheets on generators and speed regulators is considered in addition to Fig. 2.
the 𝜔 plane, as shown in Fig. 5. This study was proven by using system The dynamic structure for the frequency analysis of the microgrid
identification tools. In addition, default parameters were assigned to system was studied. The speed controller consisted of a droop and
the fractional order based on the flower pollination algorithm PID integral controller to adjust the power balance between the generator
controller [12], which will obtain the 𝐾𝑝 , 𝐾𝑖 , 𝐾𝑑 , 𝜆, 𝜇 values suitable and load. The test was performed using a virtual inertia control unit.
( )
for the plant being studied, as shown in Table 3. The parameters of the droop 𝑅𝑉 𝐼 were changed by proportionally
increasing 𝑅𝑉 𝐼 to study the virtual inertial behavior of the grid support
4. Case study mode of the microgrid system from an inverter-based ESS unit to the
transformation of the 𝑅𝑉 𝐼 image. The frequency was near the normal
In the simulation, the system is considered as a per-unit (percent) value and achieved the most appropriate response. The experimental
system, as shown in Eq. (1). The fractional-order PID controller pre- results are shown in Fig. 7.
sented in Table 3 was applied in conjunction with a virtual inertia Case 3. Effect of the virtual inertia constant.
( )
control. The power frequency efficiency was used to prove and validate A virtual inertia 𝐽𝑉 𝐼 control test was used to study the dy-
the system, owing to the penetration of renewable energy sources (wind namic behavior of the microgrid systems. The virtual inertia constant
and power plants) added to the system under various circumstances is responsible for bringing the frequency-oscillating power system to a
in various combinations with loads. This study demonstrates that the steady state until the frequency of the microgrid is stabilized. The test

6
K. Hongesombut and R. Keteruksa Electric Power Systems Research 220 (2023) 109381

Fig. 6. Effect of the virtual inertia control unit test.

Fig. 7. Effect of the virtual inertia droop control test.

Fig. 8. Effect of the virtual inertia constant test.

Fig. 9. Effect of the virtual damping test.

was performed by incrementally adjusting the simulated virtual inertia

parameters from the model presented to achieve an optimal response. Table 5
The experimental results are shown in Fig. 8. Comparison of the command-tracking performance.

Case 4. Effect of virtual Controllers type Load regulation response

( ) damping.
A virtual damping 𝐷𝑉 𝐼 control test was used to study the dynamic 𝑃 𝑂𝑟𝑒𝑔 𝑡𝑟𝑒𝑔 𝑒𝑠𝑠
behavior of the microgrid systems. It is responsible for bringing the Without 1.4030 33.5700 0.1000
frequency-oscillating power system to a steady state until the frequency PID 1.3792 32.6900 0.0000
of the microgrid system is stabilized. The test was performed by in- FOPID 0.7951 15.4200 0.0000
crementally adjusting the simulated virtual damping parameters from
the model presented to achieve an optimal response. The experimental
results are presented in Fig. 9. the communication channel, physical response time, physical settings,
Case 5. Effect of time delay.
and operation of the ESS unit using an inverter at such times, which
The delay variability test from the ESS control unit 𝑇𝐸𝑆𝑆 was used
may deteriorate the stability and efficiency of the power system. The
to study the dynamic behavior of the microgrid systems. In practice,
there is almost no response or a rapid change in the frequency of the test was performed by incrementally adjusting the delay variation
device owing to the delays and filters installed in the virtual inertia parameter from the control device of the simulated ESS unit from the
control process, considering Fig. 2. Therefore, signal measurements model presented to achieve an optimal response. The experimental
and filtering can cause delays. The delay analysis also depends on results are shown in Fig. 10.

7
K. Hongesombut and R. Keteruksa Electric Power Systems Research 220 (2023) 109381

Fig. 10. Effect of the time delay test.

Fig. 11. Effect of the controller for the virtual inertia test.

Case 6. Effect of the controller for virtual inertia. behavior. The system overshoot from the inertia (H) and damping (D)
A comparison of the traditional controls with the methods used is of 100% (blue solid line) significantly reduces the overshoot behav-
presented. The disturbed system exhibited a stable microgrid perfor- ior. This allows the system to return to a stable state and achieve a
mance. In all three cases, without the virtual inertia controller, the stable frequency as quickly as possible. This indicates that when the
PID controller and virtual inertia FOPID controller provided a longer frequency drop is more or less, depending on the MW quantity at each
settling and stabilization time as non-classical damping was simulated. generator, the capacity can be immediately increased. When the power
Using the proposed regulator, the sedimentation and stabilization time is imbalanced in a large system, significant system frequency bias
was significantly reduced, effectively improving the frequency stabi- occurs because several generators share the load to help the system.
lization rate and microgrid flexibility. The results obtained by the PID Therefore, it is necessary to properly distribute the spinning reserve
and FOPID based on the FPA are shown in Table 5, and indicated across different generators to help better normalize the system.
by (18) and (19), respectively. Meanwhile, the system was exposed to Case 2: Fig. 7 clearly demonstrates the lowest point of the fre-
renewable energy sources (RESs) disturbances and loads by switching quency from the thermal energy deviation. The power response of
the wind sources into the system at 200 s, and the solar power was the parameterized generating units of the 𝑅𝑉 𝐼 figure was found to
switched into the system at 700 s to study the dynamic behavior of be less stressful and generated less energy, as shown in the image
the microgrid system from the model presented to achieve an opti- when zoomed-in. Considering the time required to rapidly reach a
mal response. The FPA was set according to the search parameters steady state, the proposed controller performed better at the lowest
recommended by Xin-She Yang. The maximum generation (MaxGen = point. In the opposite direction, the overshoot results from the system
500) was established as the trial termination criterion (TC). Fifty trials frequency deviation. This is directly proportional to the increase in the
were conducted to determine the best solution [𝐾𝑝 , 𝐾𝑖 , 𝐾𝑑 , 𝜆, 𝜇]. For parameterization of the drawing 𝑅𝑉 𝐼 , that is, a longer time required to
comparison with the PID controller, the integration order (𝜆) and reach the steady state. In this case, an inverter-based ESS unit would
derivative order (𝜇) in Eq. (18) are set to 1.0 (𝜆 = 1.0, 𝜇 = 1.0). produce more inertial energy. Maximum frequencies were provided
Following the completion of the search procedure, the PID and FOPID while delivering longer response times to stabilize the system following
controllers optimized by the FPA were successfully obtained, as shown the disturbance from the load and infiltration of the renewable energy
in (18) and (19), respectively. Table 5 presents a comparison of the sources (wind and solar power plants).
performance, where 𝑃 𝑂𝑟𝑒𝑔 is the maximum percentage overshoot of Case 3: Fig. 8 demonstrate that the lowest point of the thermal
the load regulation, 𝑡𝑟𝑒𝑔 regulates the duration of the load regulations,
energy deviation from the microgrid system occurred when disturbed
and 𝑒𝑠𝑠 is the steady-state error. The experimental results are shown
by the renewable energy-generating units. When the 𝐽𝑉 𝐼 constant of
in Fig. 11.
magnitude is less than 1.6 (Base), it affects the system frequency at its
40.2646 lowest point until it oscillates; and in the same direction, the overshoot
𝐺𝑐 (𝑠)𝑃 𝐼𝐷 = 15.0000 + + 1.9532𝑠 (18)
𝑠 of the system from the same inertia deviation as before entering the
38.0294 steady-state is indicated by large waves. Additionally, the inverter-
𝐺𝑐 (𝑠)𝐹 𝑂𝑃 𝐼𝐷 = 13.4653 + + 2.0000𝑠0.4499 (19)
𝑠0.0365 based ESS unit produces more inertial energy to bring the system to
a steady state for a longer time. To alleviate this problem, the value
5. Results demonstration and analysis must be adjusted, and 𝐷𝑉 𝐼 is increased to achieve faster constant times.
Otherwise, the system may take a long time to reach a stable state until
Case 1: Fig. 6 that the lowest point of the frequency from a 100% a stable frequency is achieved. This type of situation is dangerous for
thermal energy deviation (blue solid line) has the longest steady-state electronic equipment used in power systems.

8
K. Hongesombut and R. Keteruksa Electric Power Systems Research 220 (2023) 109381

Case 4: Fig. 9 the system frequency response to the change in 𝐷𝑉 𝐼 interference conditions, the recommended control technique can
from an aggressive microgrid renewable energy source. The combined- efficiently maintain the frequency within a reasonable range.
cycle power unit is affected; apparently a lower 𝐷𝑉 𝐼 results in the The efficacy, flexibility, and resilience of the control techniques
system frequency bottoming out. This stabilizes the system for the can be assessed using these findings.
longest time. Compared to the constant 𝐷𝑉 𝐼 other sizes consider an 3. The study of stabilizing the microgrid frequency demonstrates
( )
additional extension (zoom); considering the system overshoot, the that a harmonized concept of virtual inertia 𝐽𝑉 𝐼 and vir-
( )
𝐷𝑉 𝐼 virtual inertia parameter is observed. When 𝐷𝑉 𝐼 increases above tual damping 𝐷𝑉 𝐼 can be achieved by building an optimum
1.2 (base), the frequency of the system is effected, leading to the regulator. The microgrid settling time minimizes the transient
pre-stabilization overload of the system and large undulations are frequency and deviation. It boosts the damping performance and
produced. This prolongs the time required for the system to adjust to a tolerates the system interference without affecting the microgrid
steady state until the frequency is stabilized. This is detrimental to the systems.
electronics used in a power system, and the inverter-based ESS unit ( )
produces more inertial energy to bring the system to a steady state for Finally, in this study, the virtual inertia 𝐽𝑉 𝐼 and virtual damp-
( )
a longer time. To solve this problem, 𝐽𝑉 𝐼 must be increased to help ing 𝐷𝑉 𝐼 were simulated in fractional-order PID controller based on
alleviate the oscillations and allow adaptation to a steady state until flower pollination algorithm, as well as the virtual inertial control for
frequency stabilization is achieved at a faster time constant. microgrid frequency stabilization. This study revealed that the pro-
Case 5: Fig. 11 presents a comparison of the frequency response posed strategy accelerates the stabilizing the frequency and maintains
of the system with the delay variation from the 𝑇𝐸𝑆𝑆 frequency re- it through severe microgrid system impacts. Future studies may apply
sponse of the thermal generator model. When working under specified FOPID to categories in other types of microgrids.
circumstances, as the 𝑇𝐸𝑆𝑆 constant increases in size, there is a greater
lift in productivity from the bottom. Simultaneously, the spike in the CRediT authorship contribution statement
frequency after the interference caused the system’s rate of change of
frequency (RoCoF) to be elevated, resulting in a reduced frequency Komsan Hongesombut: Conceptualization, Methodology, Model-
oscillation and allowing adaptation to steady state until frequency ing, Investigation, Validation, Writing – original draft, Formal anal-
stabilization was achieved at a faster time constant. ysis. Ruangyos Keteruksa: Software, Proofread, Supervision, Project
Case 6: Fig. 11 indicates that, the microgrid system was exposed administration, Investigation, Conceptualization, Methodology.
to renewable energy sources (RESs) and load disturbances. The wind
energy was switched into the system at 200 s, and the solar source was Declaration of competing interest
switched into the system at 700 s to study the dynamic behavior of the
microgrid system. This study found that the proposed fractional-order The authors declare that they have no known competing finan-
PID controller can bring the system to a steady state until the frequency cial interests or personal relationships that could have appeared to
is stabilized, thereby achieving the shortest time constant compared to influence the work reported in this paper.
the other cases, considering the zoomed image. When considering the
capabilities of the virtual inertia control technique, the experiments Data availability
in this section suggest that a system with virtual inertia control can
provide a better performance than a system without it. The system Data will be made available on request.
can achieve an optimal response in the event several disturbances; in
particular, under critical situations of low system inertia and damping Acknowledgments
caused by a high share of DGs/RESs. Systems with virtual inertia
control can mitigate the system to a steady state until the frequency The research for this project has been carried out by Financial Schol-
is stabilized, which is better than that of other methods. arship from the Royal Thai Government, under the Ministry of Higher
Education, Science, Research and Innovation (MHESI) and Faculty of
6. Conclusion Engineering at Kasetsart University.

This study introduces a fractional-order PID controller with virtual References

inertia control based on the flower pollination method. The inertia,
damping, and frequency stability of the system can be improved by si- [1] Sherif A. Zaid, et al., A new intelligent fractional-order load frequency control
( )
multaneously reproducing the virtual inertia 𝐽𝑉 𝐼 and virtual damping for interconnected modern power systems with virtual inertia control, Fractal
( ) Fract 7 (2023) 62.
𝐷𝑉 𝐼 . The following was concluded in this study:
[2] Kavita Singh, Zaheeruddin, Enhancement of frequency regulation in tidal turbine
power plant using virtual inertia from capacitive energy storage system, J. Energy
1. A FOPID controller was developed to evaluate the dynamic
( ) ( ) Storage 35 (2021) 102332.
impact of virtual damping 𝐷𝑉 𝐼 and virtual inertia 𝐽𝑉 𝐼 on [3] Li Yang, Zhijian Hu, Implementation of dynamic virtual inertia control of
the system performance. Using a large amount of virtual inertia supercapacitors for multi-area PV-based microgrid clusters, Sustainability (2020).
causes the system to accept a longer frequency deviation and [4] Ahmed M. Othman, Attia A. El-Fergany, Adaptive virtual-inertia control and
chicken swarm optimizer for frequency stability in power-grids penetrated by re-
eliminate it faster. However, increasing the virtual damping
( ) newable energy sources, in: Neural Computing and Applications, Springer-Verlag
𝐷𝑉 𝐼 may cause a large second overshoot, requiring a FOPID London Ltd., 2020, part of Springer Nature.
controller for a percent overrun of regulation to improve the [5] Thongchart Kerdphol, et al., Self-adaptive virtual inertia control-based fuzzy logic
system stability and performance. Thus, we simultaneously em- to improve frequency stability of microgrid with high renewable penetration,
( ) ( ) IEEE Access 7 (2019).
ployed the virtual inertia 𝐽𝑉 𝐼 and virtual damping 𝐷𝑉 𝐼 for
[6] Kai Shi, Haihan Ye, Wentao Song, Guanglei Zhou, Virtual inertia control strategy
stabilizing the microgrid frequency. in microgrid based on virtual synchronous generator technology, IEEE Access 6
2. Extreme RES penetration will not degrade the system. If there is (2018).
no controller with the virtual inertia, the PID controller works [7] T. Kerdphol, F.S. Rahman, Y. Mitani, Virtual inertia control application to en-
as the inertia with the virtual inertia, which is provided by the hance frequency stability of interconnected power systems with high renewable
energy penetration, Energies 11 (4) (2018) 981–997.
FOPID controller. Owing to the lack of virtual damping, fre- [8] N. Soni, S. Doolla, M.C. Chandorkar, Improvement of transient response in
quency stabilization causes instability and collapse. To prevent microgrids using virtual inertia, IEEE Trans. Power Deliv. 28 (3) (2013)
these problems, the RESs unit is disconnected. Under various 1830–1838.

9
K. Hongesombut and R. Keteruksa Electric Power Systems Research 220 (2023) 109381

[9] Tirumalasetty Chiranjeevi, et al., Control of electric machines using flower [29] Aleksei Tepljakov, Eduard Petlenkov, Juri Belikov, FOMCON: Fractional-order
pollination algorithm based fractional order PID controller, Glob. Transit. Proc. modeling and control toolbox for MATLAB, in: Conference Paper, 2011.
2 (2) (2021) 227–232. [30] C.A. Monje, Y.Q. Chen, B.M. Vinagre, D. Xue, V. Feliu, Fractional-Order Systems
[10] Emre Celik, Design of New Fractional Order PI–Fractional Order PD Cascade and Controls: Fundamentals and Applications, Springer-Verlag, London, 2010.
Controller Through Dragonfly Search Algorithm for Advanced Load Frequency [31] Javad Morsalia, et al., Comparative performance evaluation of fractional order
Control of Power Systems, Springer-Verlag GmbH, Germany, 2020, part of controllersin LFC of two-area diverse-unit power system with considering GDB
Springer Nature. and GRC effects, J. Electr. Syst. Inf. Technol. 5 (2018) 2017.
[11] F. Babaei, A. Safari, SCA based fractional-order PID controller considering [32] Xin Shen, et al., Integrated control strategy for electrolytic aluminum load
delayed EV aggregators, J. Oper. Autom. Power Eng. 8 (1) (2020) 75–85. participation in frequency modulation, IEEE Access 9 (2021).
[12] Thitipong Niyomsat, Deacha Puangdownreong, Fractional-order PID controller [33] S. Vachirasricirikul, I. Ngamroo, Robust LFC in a smart grid with wind power
design optimization for load frequency control via flower pollination algorithm, penetration by coordinated V2G control and frequency controller, IEEE Trans.
in: AICCC 2019, December 21–23, 2019, Kobe, Japan. Smart Grid 5 (1) (2014) 371–380.
[13] Lakhdar Chaib, Abdelghani Choucha, Salem Arif, Optimal design and tuning of [34] Andrew Xavier Raj Irudayaraj, et al., A Matignon’s theorem based stability
novel fractional order PID power system stabilizer using a new metaheuristic bat analysis of hybrid power system for automatic load frequency control using atom
algorithm, Ain Shams Eng. J. 8 (2017) 113–125. search optimized FOPID controller, IEEE Access 4 (2016).
[14] S.S. Mohamed, Ahmed Elbioumey Mansour, M.A. Abdel Ghany, Design of
fractional order PID controller for SMIB power system with UPFC tuned by multi-
objectives genetic algorithm, in: 16th International Conference on Aerospace
Sciences & Aviation Technology, ASAT-16-May 26-28, 2015. Komsan Hongesombut obtained the B.Eng., and M.Eng.
[15] I. Podlubny, Fractional-Order Systems and Fractional-Order Controllers, degrees in Electrical Engineering from the King Mongkut’s
UEF-03-94, Slovak Academy of Sciences, Kosice, 1994. Institute of Technology (KMITL), Bangkok, Thailand in
[16] Bo Long, et al., Frequency stability enhancement of an islanded microgrid: A 1997 and 1999, respectively. He received a PhD degree in
fractional-order virtual synchronous generator, Int. J. Electr. Power Energy Syst. Electrical and Electronic Engineering from Osaka University,
147 (2023) 108896. Osaka, Japan in 2003. From 2003–2005, he was awarded
[17] E. Rakhshani, D. Remon, A.M. Cantarellas, J.M. Garcia, P. Rodriguez, Virtual a post-doctoral fellowship by the Japan Society for the
synchronous power strategy for multiple HVDC interconnections of multi-area Promotion of Science (JSPS). As a postdoctoral fellow, he
AGC power systems, IEEE Trans. Power Syst. 32 (3) (2017) 1665–1677. worked on power system monitoring by application of GPS
[18] D. Li, Q. Zhu, S. Lin, X.Y. Bian, A self-adaptive inertia and damping combination synchronized by phasor measurement units (PMUs) in the
control of VSG to support frequency stability, IEEE Trans. Energy Convers. 32 Department of Electrical Engineering at the Kyushu Institute
(1) (2017) 397–398. of Technology, Japan. From 2005–2009, he was a specialist
[19] P. Rodriguez, E. Rakhshani, A. Mir Cantarellas, D. Remon, Analysis of deriva- in power systems at the R&D Center of Tokyo Electric
tive control based virtual inertia in multi-area high-voltage direct current Power Company, Japan. Since 2010, he has been with the
interconnected power systems, IET Gener. Transm. Distrib. 10 (6) (2016) Department of Electrical Engineering, Kasetsart University,
1458–1469. Bangkok, Thailand, where he is currently working as an
[20] S. D’Arco, J.A. Suul, O.B. Fosso, A virtual synchronous machine implementation Associate Professor. His research interest is in the area of
for distributed control of power converters in SmartGrids, Electr. Power Syst. power system analysis and protection.
Res. 122 (1) (2015) 180–197.
[21] J. Alipoor, Y. Miura, T. Ise, Power system stabilization using virtual synchronous
Ruangyos Keteruksa obtained the B.Eng. degrees in Elec-
generator with alternating moment of inertia, IEEE J. Emerg. Sel. Top. Power
trical Engineering from South-East Asia University (SAU),
Electron. 3 (2) (2015) 451–458.
Bangkok, Thailand, in 2000 and M.Eng. degrees in Instru-
[22] H. Bevrani, T. Ise, Y. Miura, Virtual synchronous generators: a survey and
mentation Engineering from King Mongkut’s Institute of
new perspectives, Int. J. Electr. Power Energy Syst. 54 (2014) 244–254, https:
Technology (KMITL), Bangkok, Thailand in 2005. He is
//link.springer.com/content/pdf/10.1007/978-0-387-84878-5.pdf.
currently pursuing the Doctor of Engineering Program in
[23] T. Shintai, Y. Miura, T. Ise, Oscillation damping of a distributed generator using a
Electrical Engineering, Kasetsart University, Thailand. His
virtual synchronous generator, IEEE Trans. Power Deliv. 29 (2) (2014) 668–676.
research interests include the stability and robust control
[24] M.A. Torres L., L.A.C. Lopes, L.A. Moran T., J.R. Espinoza C., Self-tuning virtual
synthesis of uncertainly, transient stability analysis of power
synchronous machine: A control strategy for energy storage systems to support
system and optimization of safety instrumented system.
dynamic frequency control, IEEE Trans. Energy Convers. 29 (4) (2014) 833–840.
[25] M.F.M. Arani, E.F. El-Saadany, Implementing virtual inertia in DFIG-based wind
power generation, IEEE Trans. Power Syst. 28 (2) (2013) 1373–1384.
[26] Y. Hirase, K. Abe, K. Sugimoto, Y. Shindo, A grid-connected inverter with
virtual synchronous generator model of algebraic type, Electr. Eng. Japan (2013)
(English Transl. Denki Gakkai Ronbunshi).
[27] J. Driesen, K. Visscher, Virtual synchronous generators, in: IEEE Power and
Energy Society 2008 General Meeting: Conversion and Delivery of Electrical
Energy in the 21st Century, PES, 2008.
[28] Rahmat Khezri, et al., Coordination of heat pumps, electric vehicles and AGC for
efficient LFC in a smart hybrid power system via SCA-based optimized FOPID
controllers, Energies (2018).

10