Optimal Power Flow using Advanced Optimization

Techniques

MS Thesis
by

Muhammad Ilyas

CIIT/FA19-REE-044/ISB

COMSATS University Islamabad

Pakistan

Spring 2023
Optimal Power Flow using Advanced Optimization
Techniques

A Thesis submitted to
COMSATS University Islamabad

In partial fulfillment
of the requirement for the degree of

Master of Science in
Electrical Engineering
by
Muhammad Ilyas
CIIT/FA19-REE-044/ISB
Department of Electrical and Computer Engineering

COMSATS University Islamabad

Pakistan
Spring 2023

ii
 Optimal Power Flow using Advanced Optimization
Techniques

This thesis is submitted to the department of Electrical and Computer Engineering in

partial fulfillment of the requirements for the award of the degree of Master of Science in
Electrical Engineering.

Name Registration Number

Muhammad Ilyas CIIT/FA19-REE-044/ISB

Supervisory Committee
Supervisor Co-Supervisor
Prof. Dr. Shahid A. Khan Dr. Kamran Hafeez
Professor Assistant Professor
Department of Electrical and Department of Electrical and
Computer Engineering, Computer Engineering,
COMSATS University Islamabad COMSATS University Islamabad
(CUI) (CUI)
Member Member
Dr. Junaid Ikram Dr. Obaid Ur Rehman
Associate Professor Assistant Professor
Department of Electrical and Department of Electrical and
Computer Engineering, Computer Engineering,
COMSATS University Islamabad COMSATS University Islamabad
(CUI) (CUI)
Member
Dr. Abrar Ahmed
Assistant Professor
Department of Electrical and
Computer Engineering,
COMSATS University Islamabad
(CUI)

iii
 Certificate of Approval
This thesis titled

Optimal Power Flow using Advanced Optimization

Techniques
By
Muhammad Ilyas
CIIT/FA19-REE-044/ISB

Has been approved

For the COMSATS University Islamabad, Islamabad Campus

External Examiner: _________________________________________________

Dr. Fiaz Ahmad
Associate Professor, Department of Electrical Engineering,
Air University, Islamabad

Supervisor: ________________________________________________________
Dr. Shahid A. Khan
Professor, Department of Electrical and Computer Engineering, Islamabad

Head of Department: _________________________________________________

Dr. Shurjeel Wyne

Professor, Department of Electrical and Computer Engineering, Islamabad

iv
 Author’s Declaration

I Muhammad Ilyas CIIT/FA19-REE-044/ISB hereby declare that I have produced the

work presented in this thesis, during the scheduled period of study. I also declare that I
have not taken any material from any source except referred to wherever due that
amount of plagiarism is within acceptable range. If a violation of HEC rules on research
has occurred in this thesis, I shall be liable to punishable action under the plagiarism
rules of the HEC.

Date: _21-7-2023__ Signature of the student:

____________________________
Muhammad Ilyas
FA19-REE-044

v
 Certificate

It is certified that Muhammad Ilyas CIIT/FA19-REE-044/ISB has carried out all the work
related to this thesis under my supervision at the Department of Electrical and Computer
engineering, COMSATS University Islamabad and the work fulfils the requirement for
the award of MS degree.

Date: _21-7-2023____ Supervisor:

___________________________
Prof. Dr. Shahid A. Khan
Professor
Electrical and Computer Engineering
COMSATS University Islamabad

vi
 Dedication

I would like to dedicate this work to my dearest parents, my

brother and sisters, my supervisor Prof. Dr. Shahid A. Khan and co-supervisor Dr.
Kamran Hafeez (Assistant professor) who are the relentless sources of inspiration and
knowledge.

vii
 Acknowledgements

I am forever grateful to Almighty Allah, the Creator and the Guardian of the universe,
and to whom I owe my very existence. I owe my deepest gratitude to my supervisor Prof.
Dr. Shahid A. Khan and co-supervisor Dr. Kamran Hafeez for providing me with an
opportunity to work under his supervision and supporting me throughout my thesis. He
gave me great confidence by showing strong trust and faith in me. His motivation and
encouragement helped me to work harder and smarter to explore my research. This thesis
would not have been possible unless it was for patience, motivation, enthusiasm and his
immense knowledge of the subject matter.

I would like to express my deepest gratitude to senior Manzoor Ahmad for his invaluable
support and guidance throughout the course of my research. His expertise in the field has
been instrumental in shaping my work and providing valuable insights. I am truly
thankful for his unwavering dedication and willingness to share his knowledge and
experience. His constructive feedback and suggestions have greatly enhanced the quality
of my research.

Last but not least, I am grateful to my parents, my family and friends for encouraging me
throughout my life.

Muhammad Ilyas
CIIT/FA19-REE-044/ISB

viii
 Abstract

Optimal Power Flow using Advanced Optimization Techniques

By
Muhammad Ilyas
The optimum power flow (OPF) issues in conventional thermal, solar, and wind energy
sources-based power systems have received a lot of attention recently. Finding the best
solutions to the OPF issues in the power system has been offered using a variety of
metaheuristic algorithms. The OPF is a nonlinear, nonconvex, and quadratic optimization
problem as a result of the quadratic character of its primary objective function. Because it
performs well when used to optimize the well-known Restrain quadratic benchmark
function, we have presented the Strawberry Algorithm (SBA) method in this work to
discover an optimal solution to the OPF problem in the power system. In order to achieve
precision in system operations and planning, the hybrid power system in this work
incorporates stochastic electricity generation from RESs like uncertainty wind and solar
and fixed utility load demand. We tested and measured SBA's performance using a
modified IEEE-30 bus test system, and we compared its results with well-known
evolutionary metaheuristic algorithms i-e PSO. In comparison to previous metaheuristic
algorithms, the suggested SBA consistently produces results that are more accurate,
minimize, very fast and stable. Results from simulation based optimization have
demonstrated the advantage of the SBA technique for solving OPF issues by meeting all
constraints and achieving the cost of power generation of 772.1729 $/h.

ix
 Table of Contents
Chapter 1 Introduction............................................................................................................ 1

1.1 Introduction: ................................................................................................................. 2

1.2 Optimal Power Flow (OPF): ........................................................................................ 3

1.3 Overview of hybrid ac/dc: ............................................................................................ 5

1.3.1 DC microgrid: ........................................................................................................ 6

1.3.2 AC microgrid: ........................................................................................................ 6

1.4 Strawberry Plant in Nature: ............................................................................................. 7

1.5 Particle Swarm Optimization (PSO):............................................................................. 11

1.6 Statement of the Problem:............................................................................................... 14

1.7 Research Objectives: ....................................................................................................... 14

1.8 Motivation: ....................................................................................................................... 15

1.9 Project Idea: ..................................................................................................................... 15

1.10 Organization of Thesis: ................................................................................................. 15

Chapter 2 Literature Review ................................................................................................ 17

2.1 Literature Review: ........................................................................................................... 18

Chapter 3 Uncertainty in Solar, Wind and Fixed Utility Load Demand.......................... 21

3.1 Uncertainty Modelling of Wind Speed: ......................................................................... 22

3.2 Uncertainty Modeling of Solar Irradiance: ................................................................... 26

3.3 The OPF Problem Formulation: .................................................................................... 32

3.4 Constraints: ...................................................................................................................... 33

3.4.1 Equality Constraints:............................................................................................ 33

3.4.2 Inequality Constraints: ......................................................................................... 34

Chapter 4 Methodology ......................................................................................................... 35

4.1 Thermal Power Cost Curve: ........................................................................................... 37

x
4.2 The Cost of Wind Power Function:................................................................................ 38

4.3 Function of Solar Power Cost: ........................................................................................ 40

4.4: Strawberry Algorithm Steps: ........................................................................................ 42

4.5: Particle Swarm Optimization Steps:............................................................................. 45

Chapter 5 Results and Discussion ........................................................................................ 48

5.1 Generation Cost Minimization: ...................................................................................... 53

Chapter 6 Conclusion ............................................................................................................ 55

6.1 Conclusion: ....................................................................................................................... 56

6.2 Future Works: .................................................................................................................. 57

References:.............................................................................................................................. 58

xi
 List of Figures
Figure 1: (a) Strawberry plant, (b) Desmoschoenus spiralis, (c) Chloris gayana, (d) spider plant. 9
Figure 2: Weibull fitting of wind speed ......................................................................................... 23
Figure 3: Windfarm Power (MW) ................................................................................................. 25
Figure 4: Lognormal fitting of solar irradiance. ............................................................................ 28
Figure 5: Solar PV power output. .................................................................................................. 30
Figure 6: Modified IEEE-30 bus test system. ................................................................................ 31
Figure 7: Flow chart for SBA based OPF ...................................................................................... 44
Figure 8: Flow chart for PSO based OPF ...................................................................................... 47
Figure 9: PSO Optimal Power Generation Cost Convergence. ..................................................... 51
Figure 10: SBA Optimal Power Generation Cost Convergence. ................................................... 52

List of Tables

Table 1: Different Power Systems Buses......................................................................................... 5

Table 2: Summary of Studies to Optimal solutions to the OPF problems. .................................... 20
Table 3: Windfarms and Weibull PDF parameters. ....................................................................... 23
Table 4: Solar PV units and Lognormal PDF parameters ............................................................. 26
Table 5: Thermal power cost coefficients. ..................................................................................... 38
Table 6: Cost coefficients and wind-scheduled power. ................................................................. 40
Table 7: Cost coefficients and Solar-scheduled power. ................................................................. 42
Table 8: Simulation results for electricity generation cost case studies ......................................... 49
Table 9: Case study: Trade-off ...................................................................................................... 50
Table 10: The Modified IEEE-30 Bus Test System. ..................................................................... 53

xii
 List of Abbreviations
GA Genetic algorithm
HSA Harmony search algorithm
OPF Optimal Power Flow
PDF Probability density function
PEM Point estimate method
PSO Particle swarm optimization
SBA Strawberry Algorithm
PV Photovoltaic
REEs Renewable energy sources
SHADE Success history-based adaption for DE
TG Thermal generator
WF Windfarm
WT Wind turbine

xiii
Chapter 1 Introduction

1
1.1 Introduction:

The usage of integrated RESs in existing microgrids is unavoidable due to the increasing
need for renewable energy sources (RESs), the depletion of fossil fuel resources, the need
to safeguard the environment, and the significance of reducing pollution brought on by
fossil fuel emissions. From an operational standpoint, a hybrid AC/DC microgrid is a
gateway that may improve the operational qualities of any proposed power system while
also improving the system performance in the aforementioned problems. Since the
market for renewable energy is expanding more quickly than ever, it is anticipated that
microgrids would take operational factors into account in a manner similar to how power
systems do. One result of using both AC and DC microgrids is supporting local energy
needs [1] and coupling AC and DC loads with their respective resources to reduce the
common AC-DC-AC conversion losses [2]. By combining AC and DC microgrids to
create a hybrid AC/DC microgrid, it is also possible to boost energy consumption
standards, dependability, power quality, and system stability [3,4]. Many of the factors
that contributed to the selection of a whole AC power system have mostly disappeared
after about a century of dominance [5]. The use of DC microgrids is now feasible because
to rising DC demands, semiconductor development, and increased RES accessibility in
distant and rural locations [6]. The distribution of RESs and the investments made by the
various countries in the development of hybrid AC/DC microgrids are obviously uneven.
The overall capacity/generation of the top nations as of the end of 2017 has been
examined in the worldwide status report of renewables for 2018 [7]. As can be
anticipated, the volume and quality of the energy generated and its variance are
significantly influenced by geographical variables in addition to a nation's energy
regulations. The dependence on nonrenewable resources decreases as RES penetration
increases.

According to a recent study, the typical Indian consumes 1075 kWh of energy each year,
with fossil fuels responsible for 85 percent of electricity generation, resulting in
significant volumes of CO2 pollution that lead to global warming. Renewable energy
options, on the other side, are becoming a skilled option, cost-free, environmentally
sustainable, yet with lower operational and maintenance costs, as a consequence of

2
growing demand for fuel, inadequate supplies, and the value of conventional outlets such
as firewood and coal, etc. The expanded usage of renewable energy technology Solar,
hydro, tidal, biomass, and geothermal power all have an effect on distribution generation
(DG). The introduction of DGs on a wide scale will bring organizational conflicts into the
power system network, and a microgrid with a high degree of concentration across the
world is a crucial response to this problem[8]. AC Microgrids, DC Microgrids, and
HACDC Microgrids are the three forms of microgrids [9]. Currently, the use of various
renewables is attributed to the AC Microgrids, DC Microgrids, and HACDC Microgrids
are examples of microgrids that generate both AC and DC electrical resources. DC
generation sources, such as PV, are converted to AC using DC/AC converters in AC
microgrids, whereas AC generating sources are connected directly using power electronic
interfaces [10]. DC microgrids, on the other hand, use AC/DC converters to convert AC
generating sources to DC. Many conversions are possible, on the other hand, end in
losses. The hybrid microgrid, which eliminates losses and minimizes several conversions,
is an urgent fix for the above issues. The two influential and main components Electrical
devices are framed together during hybrid systems strength and unit stability. The best
practicable system for the blueprint should be both expense and practical reliable, which
can be achieved through the correct choice of system device in the rally round. As a
result, an optimum sizing technique is needed to suggest a competent and profitable
Hybrid microgrid scheme.

1.2 Optimal Power Flow (OPF):

An important issue in electrical power systems is optimal power flow (OPF). The goal is
to reduce the cost of producing energy in an electrical power system while fulfilling
operational restrictions including power balance, voltage limits, and transmission line
capacity limits.

Finding the ideal generator setting values that minimise the overall cost of generating
while ensuring that the power system functions within its constraints is the goal of OPF.
Real and reactive power production levels, voltage magnitude, and voltage phase angle
are all parameters for the generator.

3
OPF's optimisation issue is difficult to solve since it is nonlinear, non-convex, and very
complicated. OPF problems may be solved using a variety of strategies, including linear
programming, nonlinear programming, and evolutionary algorithms.

Numerous aspects of electrical power systems, including economic dispatch, unit

commitment, power system planning, and control, have implications for OPF. Its
significance has grown as more complex optimisation methods are needed to enable the
best and most efficient integration of renewable energy sources into electrical power
networks.

In order to optimise a particular system parameter, represented by a single goal function,

an optimum power flow (OPF) is created by solving equations that describe an electrical
power system (active and reactive power of each node) and modifying the values of the
control variables (voltages or powers) [11]. A system typically consists of independent
variables (unknown data) and state variables (unknown quantities). Depending on the
goal of the study, control variables are chosen from among any of the system's
independent variables.

The parameter that will be optimised or enhanced determines the goal function of a
power system for the OPF research; some common parameters are as follows [12]:

 The cost of generation.

 Transmission losses of active and reactive power.
 Interruption costs due to flows.
 Number of controls that need to be reprogrammed.
 Pollutant emissions from thermal generators must all be kept to a minimum.

In recent years, electrical microgrids have emerged as a way to enhance power system
capabilities and lower line losses. Electrical microgrids are compact electrical systems
that have the potential for autonomous control of energy production and distribution.
They may operate in a fashion that depends on or is independent from the electrical grid
and may also be equipped with renewable energy sources [13].

4
Authors as far back as [14] - [18] have included OPF in a variety of mathematical
techniques to improve characteristics and achieve optimum performance in electrical
power systems.

Each bus has four different values: actual power (P), reactive power (Q), and voltage
magnitude (V). Buses are divided into three bus kinds in a power flow solution: PQ, PV,
and slack. In general, PV buses and PQ buses represent loads and generators,
respectively. Because power and voltage magnitude are fixed, generator buses are known
as PV buses, while load buses are known as PQ buses because real and reactive power
are fixed. The magnitude and voltage angle of slack or reference buses are constant.

Table 1: Different Power Systems Buses

Bus Type Fixed Quantities Variable Physical Model
Quantities
PV Real Power (P) Reactive Power (Q) Generator
Voltage Magnitude Voltage angle (δ)
(|V|)
PQ Real Power (P) Voltage Magnitude Load
Reactive Power (|V|)
(Q) Voltage angle (δ)
Slack Voltage Magnitude Real Power (P) An arbitrary
(|V|) Reactive Power (Q) generator
Voltage angle (δ)

1.3 Overview of hybrid ac/dc:

Hybrid AC/DC networks are a key technology for future electrical power systems.
One major reason is the increasing number of converter-interfaced energy sources
and loads. Direct current grids have several advantages [11] over traditional AC
systems: lower power losses, largely due to the absence of reactive power; higher
power transfer capability; and DC grids can also facilitate the connection of
asynchronous AC grids. However, AC technology is well established and is more

5
 suitable for some applications. Therefore, it is likely that AC and DC technology will
be combined via interlinking voltage source converters (VSC) to form a hybrid
AC/DC network [12]. Hybrid AC/DC networks present new challenges in terms of
frequency and voltage control [13]. In particular, the control of the interlinking
converter (ILC) is an important challenge [14] as its operation affects both the AC
frequency and the DC voltage.

1.3.1 DC microgrid:

The DC microgrid as we know it today was first developed in the 19th century by
Thomas Edison. A DC microgrid mainly consists of DC loads and resources. The
advantages of this type of microgrid could be energy storage system integration,
higher total efficiency due to less AC-DC-AC conversions and the elimination of
distributed generator (DG) synchronization. However, as the generated DC power
could not be transferred in long distances, it lost its popularity over time.
Nevertheless, the DC microgrid is finding its way back to the energy supply chain as
most of the home appliances such as TVs, printers, microwaves, etc. are DC supplied.
With advances of PVs and FCs as resources with DC output powers, utilizing DC
resources to supply DC loads makes more sense than ever. Research shows that about
30% of the generated AC power is transferred to DC power or at least passes through
a converter before being used. Another motivation for reconsideration of DC
microgrids is the advances in semiconductor technology. One of the challenges that
must be overcome in this microgrid, is the way of integrating the microgrid
configuration with the current distribution system.

1.3.2 AC microgrid:

This system has dominated the DC system for many years, due to its easy
modification of voltage levels with low frequency transformers and facilitated
handling of faults and protection. Moreover, the AC power is easy to transfer and
most of the industrial appliances need AC power supplement. AC RESs such as WTs,
tidal, biogas and wave turbines have been integrated with AC microgrids in recent
years. However, the main challenges in AC microgrid control are DG synchronization

6
 issues and reactive power control, which might increase the losses of the transmission
system. Moreover, since the AC RESs are sensitive to climatic and geographical
changes, frequency control of microgrids in which the aforementioned RESs are
utilized is a challenging task.

OPTIMIZATION TECHNIQUES:

1.4 Strawberry Plant in Nature:

Since they have muscles and a brain, animals are far better able to adapt to environmental
changes than plants because they may choose to relocate to locations with better
circumstances whenever necessary. The birds that migrate to warmer climates when the
weather turns chilly are probably the most well-known instances of this sort. However,
because of their rooted attachment to the ground, plants are unable to transfer to locations
with ideal circumstances. However, as seen in Fig. 1 (vegetative propagation), some
grasses and plants (such as strawberries) can be multiplied through the so-called runner
(or stolon). The runner is a creeping stalk that emerges from the mother (parent) plant and
is formed in the leaf axils. A new plant, known as a daughter plant, is created at the
second node of the runner, and a new runner then develops on the daughter plant to
produce yet another new daughter (see Fig. 1(a)). Runner plants first generate less roots
but later create an enormous amount of fibrous roots. Once the daughter plants have
amassed enough growth and roots, they may be detached from the mother plant and live
separate lives as new mother plants.

Since the mother and daughter plants share all of the same DNA and are in fact the same
plant, strawberry reproduction (as well as that of other related plants, as illustrated in Fig.
1) may be viewed as a form of plant mobility. If the daughter plant does not accidentally
arrive at a site in very poor condition, the mother plant usually passes away before the
daughter plant does.

7
(a)

(b)

8
 (c)

(d)

Figure 1: (a) Strawberry plant, (b) Desmoschoenus spiralis, (c) Chloris gayana, (d)
spider plant.
Figure. 1(c) shows the function of runners and roots for global and local search,
respectively.
9
From a mathematical perspective, plants with runners, like strawberries, carry out some
sort of Optimization. More specifically, these plants produce runners and roots (as well as
root hairs), to hunt both locally and globally for resources such as water and minerals.
The matched daughter plant generates more roots and runners when a runner or a root
(hair) moves to an area with higher water supply, which affects the development of the
whole plant. Roots and runners appear almost at random.

It goes without saying that in order to develop a numerical optimisation approach that is
inspired by the strawberry plant, we must first describe the behaviour of this plant using
simple but effective ideas. The following three facts are taken as a basis for an effective
model of strawberry plant behavior in this resreach:

 Runners that randomly appear on strawberry mother plants are used to spread the
plants' genetic material (global resource search).
 The development of roots and root hairs occurs at random on each strawberry
mother plant (local hunt for resources).
 Strawberry daughter plants that have access to more abundant nutrients develop
more quickly and produce more runners and roots, but runners and roots that
migrate towards scarce resources are more likely to perish.

The objective function does not have to be evaluated at the locations of mother plants, as
will be shown later. In the proposed SBA, we first create a random number of points
(computational agents) in the problem domain, each of which may be viewed of as a
mother plant. This is consistent with the earlier debate. Each mother plant then generates
one root and one runner (daughter plant) during each repeat; the first one is generated
close by, while the second is created a little further away. Runners and roots, which move
in the problem space in line with huge and small random steps, respectively, are really
supposed to make up the computational agents in SBA. Then, the points that runners and
roots refer to are evaluated for the objective function, and half of these points—which
likely have higher fitness values—are selected using, for instance, the roulette wheel or
elite selection and are used as the mother plants for the subsequent iteration, while the
other half are eliminated. Up until a predefined termination condition is met, this process
is repeated.

10
1.5 Particle Swarm Optimization (PSO):

The relatively new evolutionary method known as Particle Swarm Optimisation (PSO)
may be used to find optimal (or nearly optimal) solutions to numerical and qualitative
problems. Particle Swarm Optimisation was developed in 1995 by James Kennedy and
Russell Eberhart after earlier research with algorithms that mimicked the flocking
behaviour of several bird species.

In the simulations, birds would first wander aimlessly until gathering in flocks when one
of them flew over the roosting location. Due to the simple criteria the birds used to decide
their directions and velocities, a bird pushing away from the flock to roost would force
nearby birds to migrate towards the roost. As soon as they located the roost, these birds
would fly there, luring further birds to join them, and so on, until the whole flock had
made its choice.

Finding a roost is like finding a solution in a place where there are many possible options.
The likelihood that other birds will locate the roost is increased when a bird that has
located it cues them to travel in that direction.

The "socio-cognitive view of mind" is what is meant by this. According to the "socio-
cognitive view of mind," a particle largely learns from the success of its neighbors.

Each particle's velocity (or acceleration) is changed (or accelerated) towards its pbest and
gbest locations at each time step in the PSO concept. Separate random values are created
for acceleration towards the pbest and gbest locations, and acceleration is weighted by a
random term. PSO has been effectively used in several research and application domains
during the past few years. It has been proven that PSO produces better outcomes more
quickly and inexpensively than other approaches.

The parameters that must be selected carefully for the efficient performance of PSO
algorithm are:-

1. Both acceleration factors C1 & C2. (0-4)

2. Number of particles.
3. Inertia factor w

11
Power plants, an electrical transmission and distribution network, and commercial
customers linked by these networks make up the majority of an electric power grid or
power system. Through electricity generating units, also known as power plants, the
power producing facilities produce electrical power. From power generating facilities to
utilities or load centers, the transmission system transports the electricity via transmission
lines. Offices, homes, farms, businesses, and other surrounding structures receive energy
through distribution lines from the local electrical distribution system. Traditional electric
power grids are in responsible of generating energy and distributing it to consumers in the
residential, commercial, and industrial sectors through electrical transmission and
distribution lines. Two organizations, the independent system operator (ISO) and the
electric utility are in charge of overseeing the management and planning of the nation's
power system. To guarantee the dependability of energy production and the transmission
system in the electrical grid, the government created the ISO as a distinct organisation. A
entity that distributes electricity to consumers via distribution lines is known as an
electric utility. It does this by balancing the supply and demand of the electrical load.

The generating side of the electrical power grid is constantly in charge of operations and
planning, and power generation facilities modify their output in response to shifts in
consumer demand for energy. Power plants occasionally generate excess electricity,
which is either stored or delivered to the neighborhood using transmission lines [15].
Therefore, balancing the electrical system's load demand and energy supply is crucial
from a practical standpoint. Numerous methods have been used for this goal in research
literature. It is thought that problems with optimum power flow (OPF) in the power
system may be addressed in order to find stable and secure operating points for electricity
generating plants and their optimal hourly scheduling [16]-[19]. Carpentier first
recognised the expansion of the economic dispatch problem as the OPF issue in 1962
[20] in traditional thermal energy sources-based power systems. The OPF is one of the
well-known and widely-researched power system study subjects. It is characterised as
calculating the stable and secure operating points (levels) for electricity production plants
with an emphasis on reducing electricity generation cost in order to meet utility load
requirements in the power system. [20]. In typical thermal energy source-based power
systems, the OPF is a nonconvex, nonlinear, and quadratic problem due to the quadratic

12
nature of its primary objective function to reduce electricity producing costs. The
quadratic curve and its several modifications, including the quadratic curve, the
piecewise quadratic curve, and the valve-point loading effect quadratic curve with
forbidden operating zones, have been used to simulate the primary target function of the
OPF issue for the conventional thermal energy source. [20], [21]. In addition to the
primary goal of reducing the cost of energy production, academics have also proposed a
number of strategies for addressing OPF problems. Some of these objectives include
raising the voltage stability index, lowering voltage deviation, reducing power loss in
transmission lines, and lowering emissions and pollutants [20]–[23].

Due to the increased demand for electricity and concerns about global warming, the
integration of environmentally friendly and clean electricity production from renewable
energy sources (RESs) like wind and solar into thermal power systems has been essential
during the last ten years. The rapidly expanding integration of RESs into power systems
has the effect of accelerating the transition of power systems to a sustainable system
future. RESs that generate electricity, such wind farms and solar photovoltaic (PV) units,
are owned by private individuals. The ISO purchases planned renewable energy from
private sources in order to fulfil the rising load demand from customers. Wind energy
generation requires a stochastic wind speed at different times of the day. Similar to this,
irregular sun irradiation determines how much solar PV electricity is produced during the
day. The electricity produced by solar PV units and windfarms may be more or less than
the energy planned by the wind at various times of the day due to the variable and
intermittent nature of solar and wind power generation. When the amount of energy
generated by windfarm operators and solar PV units is less than the amount of wind-
scheduled power, the ISO is obligated to maintain a spinning reserve based on utility load
demand. To balance supply and demand in this situation, the ISO must raise the reserve
cost associated with reserve power producing units. An underestimation scenario may
arise when the actual quantity of renewable energy from RESs received exceeds the
planned amount. The surplus electricity generated by RESs is wasted and ISO is charged
a penalty price [22], [23] if it is not stored or moved to a nearby place. The inclusion of
stochastic power production from solar and wind sources raises the bar for controlling
and planning power systems. Utility load demand is also unpredictable because to

13
changes in consumer load demand, which directly affect the price of spinning reserves in
the power system. To achieve precision in the system's operations and planning, it is also
crucial to take utility load demand fixed into account. Therefore, a practical method is
needed to lower the overall cost of power generation.

1.6 Statement of the Problem:

The authors of research [22] suggested success history-based adaptation differential

evolution (SHADE) as a method for locating the best solution to optimal power flow
issues in power systems. Reducing the cost of energy generating and pollution from
emissions are two goals of the OPF issue. By taking into account goals to lower power
loss in transmission networks and the price of active power production, the author of
research [23] suggested a fuzzy logic approach based on SBA to discover the best
solution to multi objective optimal power flow issues in power systems. The power
system in both studies [22], [23] incorporates stochastic solar irradiance and wind speed
uncertainty to address OPF issues. However, in these research [22], [23], the utility load
demand has been disregarded. On the other hand, because DE has an early convergence
behaviour [24] and ABC is unable to seek around current remedies for nonlinear
quadratic optimisation issues [25], SHADE and ABC may be ineffective at finding
optimum solutions to the nonlinear and quadratic OPF issues.

The OPF is a nonlinear, nonconvex, and quadratic optimization problem as a result of the
quadratic character of its primary objective function. The uncertainty in solar irradiance
and wind speed in power systems incorporated with OPF makes it a challenging issue.

The power system study presented in literature incorporates wind speed and solar
irradiation that are random uncertainty to address OPF issues using different algorithms.

1.7 Research Objectives:

In this paper, we suggested a Strawberry method (SBA) method for finding the best
solutions to the OPF issues in the conventional hybrid power system based on thermal,
wind, and solar energy sources. The outcomes of simulation-based optimisation have

14
shown the SBA's supremacy in solving OPF issues by quickly and efficiently achieving
optimised power generating costs while meeting all limitations. The precision, speed,
stability, and efficiency of SBA's optimisation performance have made it appealing for
use in real-time optimisation issues. The outcomes of the simulation have inspired more
research.

1.8 Motivation:

The reason for picking this project is that using renewable energy sources like solar
energy, wind energy, thermal energy, and hydroelectric power in current micro grids will
make a difference because fossil fuels are depleting daily and polluting our environment
to a greater extent. Since the market for sustainable or renewable energy is expanding
faster than at any previous time, it is only natural that functional considerations, such as
those for power systems, would be given to microgrids. assisting regional energy needs.

1.9 Project Idea:

1. Idea is to basically full the growing demands of interconnected and isolated

communities.
2. A hybrid microgrid would be used which will provide higher power conversion
efficiency, less energy storage capacity, and higher reliability.
3. Idea is basically to use renewable energy resources rather than fossil fuels.
4. To make the environment less polluted than it is.

1.10 Organization of Thesis:

This thesis work is categorized here one by one chapter wise.

 In chapter 1, Explain a brief and basic review about Optimal Power Flow in
hybrid microgrid, Effect on environment by using renewable sources of Energy.
Optimization Techniques, Research objectives and statement of the problem are
also present in this chapter.

15
 In chapter 2, a proper Literature review is Explained. It shows a lot of efforts by
different researchers on Optimal Power Flow in power systems by different
techniques.
 In chapter 3, Explain Uncertainty modeling of Solar irradiation, fixed utility load
demand and wind speed, OPF problem formulation and its Constraints.
 In chapter 4, Explain Methodology, Define objective function, Thermal power
cost curve, Wind power cost function and Solar power cost function. SBA & PSO
steps and flow charts are explained
 In chapter 5, Discuss research results and its cost minimization values.
 In chapter 6. Conclusions and future works of this thesis are discussed.

16
Chapter 2 Literature Review

17
2.1 Literature Review:

Numerous research have used two alternative optimisation strategies for power system
OPF concerns, according to academic literature. The two types of optimisation
algorithms are metaheuristic algorithms and traditional mathematical methods. Numerous
mathematical optimisation techniques, including linear programming [26],
linear/quadratic programming [27], sequential linear programming [28], the Newton
method [29], generalised benders decomposition (GBD) [30], nonlinear programming
[31]–[33], mixed integer nonlinear programming (MINLP), [34], interior point method
[35], [36], and simplified gradient method [37], have been used to solve the optimal
power flow problems. Prior to addressing the OPF problem, nonlinear objective functions
and constraints are first transformed into linear form since the mathematical technique is
unable to handle the nonlinear aspects of the problem. [38]. This convergence of
constraints and goal functions may have an effect on the accuracy of the operations and
planning of the power system.

Using metaheuristic techniques, researchers have thoroughly investigated the optimum

power flow issue in thermal energy source-based power systems. In the last 10 years,
several research using metaheuristic algorithms have been published. These include the
moth swarm algorithm (MSA), the binary backtracking search algorithm (BSBA), the
adaptive group search optimisation (AGSO), the improved colliding bodies optimisation
(ICBO), the differential search algorithm (DSA), the stud krill herd (SKH), the
differential evolution (DE), the hybrid genetic algorithm (GA), and the binary
backtracking search algorithm (BSBA). The acronyms for numerous subjects and
methods are shown in Table 1. The major aim of the OPF issue, the minimization of
energy production costs, is taken into consideration in all of the aforementioned study
[20], [39]-[53]. Some studies also include other objectives like reducing air pollution and
power loss in transmission lines. In these research, one or more bus test systems
compliant with IEEE-30, IEEE-57, and IEEE-118 standards have been used to assess the
efficacy of the proposed metaheuristic algorithms. The power systems are rapidly rising
with the inclusion of RESs due to the problem of global warming and the increasing
demand for electricity produced by conventional thermal energy sources.

18
Several studies have been published in the literature [54]–[58] with the aim of
determining the best solutions for the optimum power flow problems in hybrid power
systems based on conventional thermal and wind energy sources. Best guided ABC
(GABC) has been employed in a research to address concerns with optimum power flow
in systems fueled by thermal and wind energy sources [54]. Their objectives were to
lower the cost of electricity production and emissions-related pollution. The modified
bacterium foraging algorithm (MBFA) is used in research [55] to address the optimum
power flow issue in the traditional thermal and wind energy sources-based power system.
In the study described above, the problem is formulated using a variety of objective
functions, and the OPF issue inequality requirements are justified using a doubly fed
induction generator model. The authors of the study [56] used ant colony optimisation
(ACO) and MBFA to solve OPF difficulties in traditional thermal and wind energy-based
power systems. In the work [57], the authors employed multi-objective glowworm swarm
optimisation (GWSO) to solve challenges with optimum power flow in power systems
using thermal and wind energy sources. The authors of the aforementioned study [38]-
[41] employed a modified IEEE-30 bus test system to verify and evaluate the efficacy of
the established approaches. According to the authors of research [58], self-adaptive
evolutionary programming (EP) was used to overcome OPF issues in conventional
thermal and wind energy-based hybrid power systems. The authors of the studies [54]–
[58] cited above employed the well-known Weibull probability density function (PDF) to
represent the uncertainty of stochastic wind speed in order to include wind power
production into power networks.

In the preceding 10 years, studies to find the best solution to the optimum power flow
problems in power systems based on thermal, solar, and wind energy sources have also
been documented [59]-[69]. The grey wolf optimizer (GWO) [59], fuzzy membership
function based SBA (FMF-SBA) [60], improved adaptive DE (IADE) [61], modified
imperialist competitive algorithm based on sequential quadratic programming (MICA-
SQP) [62], modified JAYA [63], hybrid of phasor SBA and GSA [64], and barnacles mat
[65] are just a few of the metaheuristic methods used in recent studies [59 Research [59]–
[69] has mostly employed Weibull PDF and Lognormal PDF to represent the stochastic
variability of wind speed and solar radiation, respectively.

19
Table 1 provides a summary of all the aforementioned studies [54]-[69] on hybrid power
systems based on thermal, wind, and solar energy sources. According to Table 1, all
studies take into account the basic aim of the OPF issue, which is to minimise the
quadratic fuel cost q (cost of power generation). Additionally, other goals like lowering
emission pollution Ep and power loss in transmission lines Ploss are taken into account in
nine and ten research, respectively. As can be seen from Table 1, the OPF issues in
hybrid power systems were often solved by including the unpredictability of the power
production from RESs.

Table 2: Summary of Studies to Optimal solutions to the OPF problems.

Objectives Energy Sources Uncertainty Models
Algorithm
ξq Ploss Thermal Solar Wind Solar Wind Utility
irradi speed Load
DE [24] Yes No Yes Yes Yes Yes Yes No
PSO [29] Yes Yes Yes Yes Yes Yes Yes No
GABC [54] Yes No Yes No Yes No Yes No
MBFA & GA [55] Yes No Yes No Yes No Yes No
ACO&MBFA [56] Yes Yes Yes No Yes No Yes No
GWSO [57] Yes Yes Yes No Yes No Yes No
Self adaptive EP [58] Yes No Yes No Yes No Yes No
GWO [59] Yes No Yes Yes Yes Yes Yes No
FMF-PSO [60] Yes Yes Yes Yes Yes Yes Yes No
IADE [[61] Yes Yes Yes Yes Yes No No No
MICA-SQP [62] Yes No Yes Yes Yes Yes Yes No
MJAYA [63] Yes Yes Yes Yes Yes No No No
Hybrid PPSO & Yes Yes Yes Yes Yes Yes Yes No
GSA [64]
BMO [65] Yes Yes Yes Yes Yes Yes Yes No
PSO [66] Yes No Yes Yes Yes Yes Yes No
Hybrid DE & PSO
Yes No Yes Yes Yes Yes Yes No
[67]
MBFA [68] Yes Yes Yes Yes Yes Yes Yes No
SFO [69] Yes No Yes Yes Yes Yes Yes No
Quadratic fuel cost (ξq), Power loss in transmission lines (Ploss).

20
Chapter 3 Uncertainty in Solar, Wind and Fixed
Utility Load Demand

21
The deployment of a suitable PDF for projecting the values of uncertain or random
variables is the most important part of uncertainty modelling. In accordance with the
methods stated in research [24], we employed the Lognormal PDF and Weibull PDF to
simulate the uncertainty of solar irradiance and stochastic wind speed. The utility load
demand on load buses has been discovered, per study [61].

The model for incorporating wind power output while controlling wind speed uncertainty
is given in this section's first paragraph. Also given is the model for incorporating
stochastic solar electricity into the power grid.

3.1 Uncertainty Modelling of Wind Speed:

The popular Weibull PDF has mostly been used in research literature to predict wind
speed v (m/s) [38]–[40] and [42] in order to include the stochastic character of adding
wind power production to the grid. The following are some definitions of the uncertainty
modelled by Weibull PDF for stochastic wind speed:

( )
( / )
for (1)

Where c: scale factor and k: shape factor are parameters, and its mean (Mwbl) may be
determined as follows:

(2)

where the gamma function Γ () is shown, and it may be written as:

(3)

22
 Figure 2: Weibull fitting of wind speed
To assess the effectiveness of SBA (windfarms), our modified IEEE-30 bus test system
substitutes wind energy sources for traditional thermal energy sources at the 5 generator
buses. Each windfarm's total number of wind turbines (WTs) and the c and k Weibull
PDF parameter values are shown in Table 2. Figure 4 shows the frequency distributions
and Weibull fitting of the wind speed of the wind farm energy sources linked to buses 5
after the execution of 8000 Monte Carlo simulation scenarios.

Table 3: Windfarms and Weibull PDF parameters.

S.No Energy Source WTs Rated PDF Weibull
Power Parameters mean
(MW)
1 WF5 25 75 C=9, k=2 V=7.953 m/s

The actual power production of a wind turbine in a wind farm relies on the wind speed v
(m/s) it encounters. Each WT in wind farms is rated to produce 3 MW of power. Wind

23
farms connected at generator buses 5 generate a total of 75 MW from 25 WTs. It is
possible to formulate the WT electricity generation as follows [24]:

for out
in
for in (4)
in
for in and out

When the rated wind speed is represented by WT, the rated power output is represented
by Pwr, and Vin, Vout, and Vr are cut-in. For the purpose of calculating wind power
output, Vr = 16 m/s, Vout = 25 m/s, and Vin = 3 m/s have been taken into consideration.
The wind farms connected at buses 5 provide wind energy according to the histograms in
Figure 5, which are based on the Weibull distribution of wind speed seen in Figure 4.
When the wind speed is between [Vr Vout], the WT generates the rated power output
Pwr, as can be seen from the WT power output Eq. 4. The calculation for the wind power
output is [24]:

24
 Figure 3: Windfarm Power (MW)

{ } . (5)

When the wind speed is between [Vin Vr], the WT delivers wind power continuously,
and the region's wind power output is [24]:

( )

25
 When the wind speed is not in the desired range, no energy is produced [Vin Vout].

in out
(7)
3.2 Uncertainty Modeling of Solar Irradiance:

The random solar energy production in power systems is often handled using the
lognormal PDF-based probabilistic model for solar irradiance I (W/m 2) [58]. In hybrid
power systems, the lognormal PDF-based probabilistic model for solar irradiance I
(W/m2) is frequently used to address stochastic solar power generation [58]. The sun
irradiation I (W/m2) uncertainty can be mathematically stated as follows:

( )
for (8)
√

where the solar irradiance mean is µ and the solar irradiance standard deviation is σ. The
definition of the lognormal PDF mean (Mlgn) is:

(9)

In Table 3, the median solar irradiation (µ) and standard deviation (σ) have been assumed
values. Two thermal generators at Buses 8 and 13 have been swapped out for two solar
energy sources in a modified IEEE-30 bus test setup.

Table 4: Solar PV units and Lognormal PDF parameters

S.No Energy Source Rated Lognormal Lognormal Mean,
Power PDF Mlgn
(MW) Parameters
1 PV8 35 µ=6, σ=0.6 I=485 w/m2
2 PV13 50 µ=6, σ=0.6 I=485 w/m2

26
The energy sources for the 8 and 13 buses were two solar PV units are linked together.
After running simulations of 8000 Monte Carlo scenarios, the lognormal PDF fitting and
solar irradiance frequency distributions for solar energy sources linked at buses 8 and 13
are created. Figure 6 displays these outcomes. The power generated by the solar PV unit
that is in use or the actual solar energy output that is affected by sun irradiation I (W/m 2)
is modelled as [24]:

for
(10)
for
×

where Istd and Ic stand for the solar irradiance and a specific solar irradiance point,
respectively, in a normal environment. Ppvr stands for the rated power output of a solar
PV unit.

27
 (a) Bus 8

(b) Bus 13

Figure 4: Lognormal fitting of solar irradiance.

28
In this investigation, Istd and Ic were set at 800 and 120 W/m 2, respectively. The solar
PV units that are linked to generator buses 8 and 13 have rated power outputs of 35 MW
and 50 MW, respectively. The distribution of solar irradiance I (W/m 2) shown in Figure 6
is used to construct the histograms in Figure 7, which show the stochastic solar power
output from solar PV units.

(a) Bus 8

29
 (b) Bus 13

Figure 5: Solar PV power output.

30
Figure 6: Modified IEEE-30 bus test system.

31
Our suggested solution, which substitutes three renewable energy sources (RESs) for
three thermal energy sources at generator buses 5, 8, and 13, in order to reduce emission
pollution and combat global warming, has been put to the test using a modified IEEE-30
bus system.

At Bus 1,2 and 11 thermal Generators are connected at Bus 1,2 and 11 G 1: 150MW, G2:
80MW, and G11: 50MW.

At Bus 5, Wind generator are connected, W5: 75MW.

At Bus 8 and Bus 11 Solar Photovoltaic Generators are connected, PV 8: 35MW and
PV13: 50MW.

The load or PQ bus is represented by the downward-pointing arrow () in Figure 6.

In order to represent the utility active load demand throughout the redesigned IEEE-30
bus test system, the distributed active load demands on load or PQ buses were established
after the execution of 8000 Monte-Carlo simulation scenarios. The base active load of
each load bus was assumed to be equal to the active (actual) load demand mean value.

3.3 The OPF Problem Formulation:

The OPF, which is a quadratic nature nonconvex and nonlinear issue, is solved in the
traditional thermal energy sources-based power system by obtaining stable and secure
operating point settings in electricity generating plants in order to minimise certain
objectives. The goal of the OPF issue is expressed as [61]:

Minimize: f (x, u)

s.t: g (x, u) = 0 (11)

h (x, u) ≤ 0,

there is both a set of state variables (x) and a set of control variables (u). The goal of the
optimal power flow issue is represented by the function f (x, u). Both equality and
inequality restrictions are represented by the functions g(x, u) and h(x, u).

32
Control or independent variables oversaw the movement of power in the system, whilst
state variables reflected the status of the power system. The control variables consist of
all active bus generators, with the exception of slack (swing) bus generators, all generator
or energy source voltage magnitudes, shunt compensator at particular buses, and
transformer tap in the power system network. The state variables are generated from
reactive power output from generators, active power output from swing buses, line
loading on transmission lines, and voltage magnitude at load buses.

3.4 Constraints:

While security restrictions and the operational limits of the equipment, such as
transmission lines and load buses, balance between active (actual) and reactive power
while adhering to equality criteria. Here is an explanation of both kinds of limitations.

3.4.1 Equality Constraints:

Active (real) power generation from energy sources must match active (real) load
needs, as well as power loss in transmission lines, in a power system. Similar to
this, the output of reactive power from all energy sources must match both
demand and reactive power loss. According to study [61], equality restrictions in
the power system may be expressed as follows:

𝑃 −𝑃 = 𝑉∑ 𝑉 𝐺 cos 𝜃 + 𝐵 sin 𝜃
∀𝑖 ∈ 𝑁𝐵 (12),
𝑄 −𝑄 = 𝑉∑ 𝑉 𝐺 sin 𝜃 − 𝐵 cos 𝜃
∀𝑖 ∈ 𝑁𝐵
(13)
where the terms PGi and QGi denote the active and reactive power production,
respectively, from the energy sources connected to the i-th bus. The terms PDi
and QDi refer to the active power load demand and the reactive power demand,
respectively, at the ith load bus. The word Vi denotes the magnitude of the ith bus
voltage magnitudes, whereas the term Vj denotes the jth bus voltage magnitudes.

33
 The expression ϴij = ϴi - ϴj demonstrates the distinction between the voltage
angles at the ith and jth buses. The terms Gij and Bij refer to transfer conductance
and susceptance, respectively.

3.4.2 Inequality Constraints:

Security restrictions on load buses and transmission lines, physically secure and
stable equipment settings, and the maximum and lower operational limitations of
equipment are all regarded as unequal constraints in power systems. The
following is how these are mathematically explained [61]:

max
, , ,

, , ,

, , ,

. , . ,

While NTG, NWF, and NPV represent the total number of each kind of energy source,
respectively, The active (actual) power output boundary limits of RESs like wind farms
and solar photovoltaic (PV) units are shown in constraints (14) through (16). Constraints
(17) through (19) define the reactive power output boundary limitations of all energy
sources or generators, including conventional thermal energy sources, wind farms, and
solar PV units. the constraints (20) and (21) that pertain to the voltage magnitude
boundary restrictions for generator buses and load buses, respectively. There are NG
generator buses. In each inequality constraint, the superscripts max and min indicate the
corresponding parameter's boundary bounds. There are NL load buses.

34
Chapter 4 Methodology

35
Lowering the cost of electricity production from traditional thermal energy sources and
RESs is our objective as we try to figure out the best solution to the OPF issue in the
power system. The goal of reducing the cost of power output is as follows:

Objective Function:

𝑵𝑾𝑭
𝑻 𝑻𝑮 𝒋 𝟏 𝒘,𝒋 𝒘,𝒋
𝑵𝑷𝑽
𝒌 𝟏 𝒑𝒗,𝒌 𝒑𝒗,𝒌
(22)

The first component, 𝝃𝑻 (𝑷𝑻𝑮 ), denotes the cost of producing electricity from thermal
energy sources using a quadratic curve to account for valve-point loading effects. The
cost of electricity generation from wind farms is represented by the second term,
∑𝑵
𝒋 𝟏 𝝃𝒘,𝒋 𝑷𝒘,𝒋 . Cost of solar power generation from solar PV units is the third
𝑾𝑭

𝑵
component ∑𝒌 𝑷𝑽𝟏 𝝃𝒑𝒗,𝒌 𝑷𝒑𝒗,𝒌 in the objective function. The following provides a
thorough description of the cost functions for traditional thermal and RESs power
generation:

PSO Parameters:

N_particles: Number of Particles

c1, c2, w: PSO hyperparameters.

Bounds: the boundary of the candidate solutions within a specific rang (min=0,
max=Ymax)

Iters: number of iterations

SBA Parameters:

Sroot, Srunner are two scalars representing the distance of roots and runners from the
mother plant respectively.

numSol: Number of solutions in the initial population.

36
Nm is the number of mother plants where 1<Nm<=Nd.

N: Number of strawberry plant generations.

4.1 Thermal Power Cost Curve:

Electricity was produced using a fossil fuel-based thermal energy source. Regular
quadratic fuel curves, quadratic fuel curves with valve-point effects, and
piecewise quadratic fuel curves may all be used to determine the cost of
producing electricity using fossil fuels [54]. In certain real-world situations,
thermal power is produced utilising conventional thermal energy sources and a
variety of fossil fuels, including coal, oil, and natural gas. The piecewise quadratic
fuel cost curve is used to determine the cost of the power production from various
kinds of energy sources. In this analysis, we made the assumption that power
generation uses the same traditional thermal energy sources that are based on
fossil fuels. Therefore, we employed two types of fuel cost curves—quadratic fuel
curve and valve-point effects quadratic fuel curve—to evaluate the power cost
associated with thermal energy source. According to [61], The amount of
electricity (MW) generated by thermal energy sources is inversely proportional to
the cost of fossil fuels per hour.

(23)

where PTGi is the output power of the ith thermal energy source. The cost
coefficients of the ith thermal energy source are ai, bi, and ci. To evaluate the cost
of thermal power in a typical thermal energy source-based power system more
accurately and realistically, the target function is simulated using valve-point
loading effects quadratic fuel cost curve. Such a circumstance leads to a deviation

37
 in the fuel cost curve since the steam turbines used in conventional thermal
energy sources have several valves. The thermal energy source's fossil fuel cost
curve is calculated in this case using the valve-point loading affects quadratic fuel
cost curve, as illustrated in [61]:

𝜉 (𝑃 ) = ∑ 𝑎 +𝑏 𝑃 +𝑐𝑃
(24)
+ 𝑑 × sin 𝑒 × 𝑃 −𝑃 ,

And the valve-point loading impact of the ith TG is represented by coefficients di

and ei. Pmin TGi denotes the operating ith TG's minimal power output. Table 4
details the cost coefficient values for the gasoline cost curve and the impacts of
valve-point loading on TG1, TG2 and TG3 for this investigation.

Table 5: Thermal power cost coefficients.

Energy Bus No. a b c d e
Source
TG1 1 0 2 0.00375 18 0.037
TG2 2 0 1.75 0.0175 16 0.038
TG3 11 0 1.75 0.0175 16 0.038

4.2 The Cost of Wind Power Function:

We've presummit that individuals own RESs like solar PV systems and wind
farms, and that ISO buys scheduled power from individuals in accordance with a
written contract. The cost of wind power is directly proportional to wind-
scheduled power and may be expressed as follows :

38
 , , , (25)

where λws,j is a function that estimates the cost of wind-generated electricity.

Pws,j is the amount of wind-scheduled electricity that was received from the jth
windfarm, and gj is the cost coefficient for that power.

Figure 5 displays the power output distributions of wind farms. Because the
production of wind farms is stochastic and variable, the electricity that is available
from them may be less or greater than the power that is planned by the wind. If
the amount of wind power generated by wind farm operators is less than the
amount of wind power planned, the ISO is required to maintain a spinning reserve
based on utility load demand. For the jth windfarm, the wind power reserve cost
may be calculated using the following formula:

,
, , , , , ,
(26)
, ,

where Pwav,j, Kwr,j, and w(Pw,j) denote the current wind energy available, the
reserve cost of the jth wind farm, and the wind energy, respectively. When wind
power supplied by windfarm operators exceeds wind-scheduled power and it is
not practicable to lower power production from thermal energy sources, the
excess electricity from the windfarms is dumped and ISO pays a penalty cost. It is
possible to construct the fine associated with the jth windfarm as [55]:

,
𝜆 , 𝑃 , −𝑃 , =𝐾 , ∫ 𝑃 , −𝑃 ,
,
(27)
×𝑓 𝑃 , 𝑑𝑃 , ,

39
 where Kwp,j is the wind farm's jth wind penalty cost coefficient. Penalty, reserve,
and wind-scheduled power costs are added to determine the cost of any wind farm
electricity. The cost coefficients and planned wind power are shown in Table 5.
The wind scheduled power cost, as well as the penalty and reserve charges, may
be added to determine the overall wind power cost ξw,j(Pw,j) for the jth windfarm
as shown in [55]:

, , , , , , ,
(28)
, , ,

Table 6: Cost coefficients and wind-scheduled power.

Energy source Scheduled Scheduled Penalty Cost Reserve Cost
Power (MW) Cost coefficient coefficient
coefficient
WF5 Pws,1=[0 75] g1=1.6 Kwp,1=1.5 Kwr,1=3

4.3 Function of Solar Power Cost:

The price of solar energy is inversely correlated with solar scheduled power. The
cost of solar-scheduled power λpvs,k depends on the scheduled power delivered
by the kth solar PV unit in the manner shown in [58]:

, , , , [29]

where Ppvs,k is the supplied solar-scheduled power from the kth solar PV unit,
and scheduled power cost coefficient hk is connected to the kth solar PV unit.
The distributions of electricity produced by solar PV units are depicted in Figure
7. Similar to the power production behaviour of wind farms, the available power
from solar PV units may be less than solar-scheduled power in an overestimation

40
 scenario and more than solar-scheduled power in an underestimation
circumstance. In such a case, the ISO need a spinning reserve energy source.
Under an overestimation scenario, we have estimated solar reserve cost (pvr,k) of
kth solar PV unit power output. in accordance with the theory outlined in the
paper [58]. It may be stated as follows [58]:

𝜆 , 𝑃 , −𝑃 , =𝐾 , ×𝑓 𝑃 , <𝑃 ,
(30)
× 𝑃 , −𝐸 𝑃 , ,< 𝑃 , ,

where Ppva,k represents the actual power received by the kth solar unit, Kpvr,k is the
coefficient of reserve cost for the kth solar PV unit, and ƒpv(Ppva,k < Ppvs,k) is the
computation of the overestimation scenario pertaining to the kth solar PV unit. The
expectation (forecast) of power production from solar PV unit kth under solar scheduled
power Ppvs,k is represented by E(Ppva,k < Ppvs,k).

The penalty cost λpvp,k associated with the power output of the kth solar PV unit
may be expressed as [58] for the underestimating scenario.

𝜆 , 𝑃 , −𝑃 , =𝐾 , ×𝑓 𝑃 , >𝑃 ,
(31)
× 𝐸 𝑃 , >𝑃 , −𝑃 , ,
Kpvp,k stands for the cost penalty coefficient for the kth solar PV unit. The
computation of underestimation situations pertaining to the kth solar PV unit is
represented by ƒpv(Ppva,k > Ppvs,k). The formula E(Ppva,k > Ppvs,k) denotes
the expectation (prediction) of power generation from a solar PV unit kth above
solar-scheduled power Ppvs,k.

Similar to how wind farm power costs are determined, the cost of a solar PV
unit's electricity is determined by combining the costs for penalties, reserves, and
solar-scheduled power. The cost coefficients and solar planned power related to
solar PV systems are listed in Table 6. By combining solar-scheduled power

41
 costs, penalty costs, and reserve costs, the total solar power cost for the kth solar
PV unit may be calculated as follows [58]:

, , , , , , ,
(32)
, , ,

Table 7: Cost coefficients and Solar-scheduled power.

Energy source Scheduled Scheduled Penalty Cost Reserve Cost

Power (MW) Cost coefficient coefficient
coefficient
PV8 Ppvs,1=[0 35] h1=1.5 Kpvp,1=1.5 Kpvr,1=3
PV13 Ppvs,2=[0 50] h2=1.5 Kpvp,2=1.5 Kpvr,2=3

4.4: Strawberry Algorithm Steps:

The various steps involved in the implementation of SBA to the OPF problem are:

Step 1: Before specifying the lower and upper bounds of each variable, read the system's
input parameters (bus, line, and generator data). A large-capacity generator is considered
at the slack bus when optimising for N generators using N-1 generators.

Step 2: Using the Newton-Raphson technique, calculate losses and the power flow
solution.

Step 3: Randomly initialize N Vectors (Mother Plant) in the domain problems.

Step 4: For each vector (mother plant), Calculate a runner and a root, and consider each
of the resulted vector as a column of (X)i

42
Step 5: Determine the fitness of the runners and roots (i.e., the fitness X(i), and then
choose N vectors (mother plants of the next iteration) from these 2N runners and roots
using elite selection and the roulette wheel. The chosen vectors are regarded as X(1) i +
columns. Set i:=i+1

Step 6: Conditions satisfied or go Back to step 2.

End

The pseudocode of SBA:

Initialize root d=10, runner r=400 , and Number of mother plants, N=50 and Number
of variables = 11
Set i:=0
Randomly initialize N vectors (mother plants) in the domain of problem. Consider
each of these random vectors as a column of X(0).
WHILE (the termination conditions are not met)
Duplication: Calculate a runner (daughter plant) and a root from (2) for each
vector (mother plant), and then use each of the resulting vectors as
a column of Xprop (i).
Elimination: Determine the fitness of the runners and roots (i.e., the fitness of
the columns of Xprop (i) from (5)), and then choose N vectors
(mother plants of the next iteration) from these 2N runners and
roots using elite selection and the roulette wheel. The chosen
vectors are taken into account as columns of X(i + 1).
Set i:=i+1
END WHILE

43
Figure 7: Flow chart for SBA based OPF
44
4.5: Particle Swarm Optimization Steps:

The various steps involved in the implementation of PSO to the OPF problem are:

Step 1: Read the system's input parameters (bus, line, and generator data) before defining
the lower and higher limits of each variable. For N generators, optimisation is done for
N-1 generators, and a large-capacity generator is taken into account at the slack bus.

Step 2; The population's particles are then randomly initialised, or are chosen at random
from the corresponding minimum and maximum values. Additionally, initialise the
velocity V between [-1 and 1].

Step 3: Using the Newton-Raphson technique, calculate losses and the power flow
solution.
Step 4: pBest is given to the fitness that is the best. At this point, the pBest and gBest are
the same.
Step 5: Iteration 𝑖 = i + 1 is updated.
Step 6: Update the inertia weight 𝑤 given by

𝑊 = −(𝑊 −𝑊 )/ iter = iter

Step 7: Adjust each particle's velocity v in accordance with the aforementioned equation.

𝑉(𝑘, 𝑗, 𝑖 + 1) − 𝑤 ∗ 𝑉(𝑘, 𝑗, 𝑖) + 𝐶1∗ rand* (𝑝𝑏𝑒𝑠𝑡𝑥(𝑗, 𝑘) − 𝑥(𝑘, 𝑗, 𝑖)) +

𝐶2∗ rand* (𝑔𝑏𝑒𝑠𝑡𝑥(𝑘) − 𝑥(𝑘, 𝑗, 𝑖))

Step 8: Each particle's position is altered in accordance with the aforementioned

equation.

A particle's location is reset to the appropriate limit if it exceeds its position restrictions in
any dimension.

𝑥(𝑘, 𝑗, 𝑖 + 1) = 𝑥(𝑘, 𝑗 − 1, 𝑖) + 𝑣(𝑘, 𝑗, 𝑓)

Step 9: Each particle is assessed in accordance with its most recent location by
computing the fitness function and running power flow. The current value is assigned to

45
be pBest if each particle's evaluation value is better than the previous pBest. The value is
set to gbest if the best p Best is superior than gBest.

Step 10: We proceed to Step 11 if one of the halting conditions is met. If not, we go on to
Step 5.

Step 11: The most recent ideal value to be produced by the particle is called gBest.

46
Figure 8: Flow chart for PSO based OPF
47
Chapter 5 Results and Discussion

48
The SBA’s simulation-based optimization outcomes are evaluated and contrasted with
PSO algorithms:

Table 8: Simulation results for electricity generation cost case studies

Parameters Min 𝐌𝐚𝐱 PSO SBA

𝑷𝑻𝑮𝟏 (𝑴𝑾) 50 150 134.93632 134.96223

𝑷𝑻𝑮𝟐 (𝑴𝑾) 20 80 37.0054 38.6058

𝑷𝑾𝟓 (𝑴𝑾) 0 75 36.8805 34.3924

𝑷𝑻𝑮𝟏𝟏 (𝑴𝑾) 15 50 36.9080 33.6343

𝑷𝑷𝑽𝟖 (𝑴𝑾) 0 35 23.2596 22.9316

𝑷𝒑𝒗𝟏𝟑 (𝑴𝑾) 0 50 20.0000 24.2730

𝑸𝑻𝑮𝟏 (𝑴𝑽𝑨𝒓) -20 150 -13.0058 -20.1506

𝑸𝑻𝑮𝟐 (𝑴𝑽𝑨𝒓) -20 60 8.0759 30.83961

𝑸𝑾𝟓 (𝑴𝑽𝑨𝒓) -30 35 24.97633 34.97633

𝑸𝑻𝑮𝟏𝟏 (𝑴𝑽𝑨𝒓) -15 40 15.6987 26.8985

𝑸𝑷𝑽𝟖 (𝑴𝑽𝑨𝒓) -20 25 18.94608 21.86608

𝑸𝒑𝒗𝟏𝟑 (𝑴𝑽𝑨𝒓) -25 30 22.28046 21.31046

𝑽𝟏 (𝒑. 𝒖) 0.95 1.1 1.1 1.1

𝑽𝟐 (𝒑. 𝒖) 0.95 1.1 1.0970 1

49
 𝑽𝟓 (𝒑 ⋅ 𝒖) 0.95 1.1 1.000 1

𝑽𝟖 (𝒑 ⋅ 𝒖) 0.95 1.1 1.0872 1.1

𝑽𝟏𝟏 ( p.u) 0.95 1.1 1.0255 1

𝑽𝟏𝟑 (𝒑. 𝒖) 0.95 1.1 1.0865 1

𝑷loss (𝑴𝑾) - - 5.3500MW 5.3994MW

Total cost
- - 774.3372 772.1729
($/𝐡𝐫)

Table 9: Case study: Trade-off

Algorithms Generation Convergence No. of
Cost ($/h) Time Generation
(Seconds)
PSO 774.3372 686.695 1000
SBA 772.1729 560 1000

Power output cost is shown in boldface in Table 8, 9 along with other optimal values of
the objective function via SBA and PSO algorithms. SBA technique achieved a fast value
of power generating cost of 772.1729 $/h as described in Table 8, 9 and it performed
better than other metaheuristic algorithms. Figure 9,10 shows a visual display of the
suggested method's optimal power generation cost convergence using PSO and SBA
algorithms. All load or PQ bus voltage magnitude profiles gathered during the
performance assessment of SBA and PSO algorithms are shown in table 8. The
recommended method, which is based on the SBA and PSO algorithm, trades off the
optimum power generation cost and convergence time in Table 9.

50
Figure 9: PSO Optimal Power Generation Cost Convergence.

51
 Figure 10: SBA Optimal Power Generation Cost Convergence.

To monitor and assess the performance SBA for finding the best solution to the OPF
issue in the hybrid power system, we customised the IEEE-30 bus test system. The
superiority of the SBA technique and fast convergence speed to solving OPF issues has
been shown by simulation-based optimisation results (see Table 8,9 and Figure 9, 10).

52
Table 10: The Modified IEEE-30 Bus Test System.
Variables/Items Quantity Details
Buses 30 [64]
Branches 41 [64]
TG (1,2,11) 3 Thermal Energy sources connected at bus 1
(swing), 2 and 11
Windfarms (WF5) 1 Wind Energy sources connected at bus 5
PV Units (PV8, PV13) 2 Solar Energy sources connected at bus 8
and 13
Control Variables 11 Active (real) power of generators bus: TG2,
TG11 WF5, PV8 and PV13: all generators’
buses voltage magnitudes.
Load Buses 24 In rang [0.95 – 1.05] p.u voltage magnitude
bases on 100 MVA

We have utilised the same parameter values and iterations for the SBA method and PSO
algorithms to ensure a fair comparison. Generator bus voltage upper and lower limits
have been set at [0.95, 1.1] p.u. based on 100 MVA for experimental reasons. The load or
PQ buses' voltage magnitude boundary restrictions have been maintained at [0.95 1.05]
p.u. For performance evaluation, the modified IEEE-30 bus test system with a 285 MW
power generating capacity was used. Further information is provided in Table 8.

We utilized MATLAB R2022a to implement the suggested approach and other methods,
using the MATPOWER6.0 package to calculate load flow. On Intel Core(TM) i7-2600U
CPUs running at 3.40 GHz and 8.00 GB of RAM under Microsoft Windows 10 64-bits,
simulation-based studies have been conducted. In order to assess the effectiveness of the
suggested BSA technique, we performed two case studies. In order to lower the cost of
energy generation, we first addressed the optimal power flow issues in a power system.

5.1 Generation Cost Minimization:

For the purpose of evaluating the performance of the suggested SBA-based
technique and other metaheuristic algorithms to address OPF issues in a hybrid
power system, simulation-based tests have been carried out. In Eq. 22, the OPF
problem aim to lower the cost of power output is expressed. The first component,
𝝃𝑻 (𝑷𝑻𝑮 ), in Eq. 22 denotes the cost of power production associated with thermal

53
energy sources. The cost of electricity output connected to wind farms is
𝑵
represented by the second phrase, ∑𝒋 𝑾𝑭
𝟏 𝝃𝒘,𝒋 𝑷𝒘,𝒋 ., while the cost of solar PV
𝑵
units is represented by the third term, ∑𝒌 𝑷𝑽𝟏 𝝃𝒑𝒗,𝒌 𝑷𝒑𝒗,𝒌 .

Three different forms of wind-related costs, such as 1) the cost of wind-scheduled

power, 2) the cost of reserves, and 3) the cost of penalties, are added to determine
the cost of an individual wind farm's energy production. The cost of wind-
scheduled power for each particular wind farm is directly related to wind-
scheduled power, and when wind-scheduled power is kept high, a high spinning
reserve is needed. In such a scenario, the cost of wind electricity goes up overall
while penalty costs go down. The scale parameter c of the Weibull PDF and the
lowest wind power cost attained at an intermediate value have a significant impact
on wind speed and wind electricity output. Similar to this, the cost of producing
energy from a single solar PV unit is determined by combining the costs
associated with penalties, reserves, and solar-scheduled power. A clear correlation
exists between the price of solar-scheduled power and solar PV unit costs. It has
been noted that the cost of solar power generation did not grow monotonically
with the values of lognormal PDF parameters like the mean and standard
deviation for solar irradiation. Consequently, choosing the appropriate value of
solar-scheduled power linked with solar PV units requires careful consideration. It
is advised to select a lower number for solar-scheduled power if the mean value is
kept low.

54
Chapter 6 Conclusion

55
6.1 Conclusion:

In this work, we have provided a new efficient value for the OPF issues in the
conventional thermal, wind, and solar energy sources-based hybrid power system using
the Strawberry method (SBA) method. This includes utility load demand and the
stochastic nature of RES power supply. An equation for the valve-point loading affects
quadratic fuel curve is used to calculate the cost of producing electricity from thermal
energy sources. Weibull PDF and Lognormal PDF have each been used for stochastic
solar irradiance and wind speed, respectively. The findings of simulation-based
optimisation have shown the superiority of the SBA in resolving OPF concerns by
satisfying all requirements, and in the example study, quick power generation at a
minimum cost of 772.1729 $/h was achieved. According to optimisation results, the
output of clean energy from RESs has increased while the hybrid power system's
emission pollution has decreased. The SBA approach is better than PSO metaheuristic
algorithms, according to the results of the comparative analysis and simulation-based
optimisation. SBA has an advantage over other optimisation techniques in terms of cost
minimization, precision, stability, and efficiency, which makes it appealing for use in
real-time optimisation issues. The outcomes of the simulation have inspired more
research.

56
6.2 Future Works:

Future applications of the suggested SBA technique might include solving additional
optimisation issues including unit commitment, chance-constrained OPF, and transient
stability restricted OPF in a large-scale conventional thermal energy sources-based power
system.

57
 References:
[1] Karabiber A, Keles C, Kaygusuz A, Alagoz BB. An approach for the integration of
renewable distributed generation in hybrid DC/AC microgrids. Renew Energy
2013;52:251–9. https://doi.org/10.1016/j.renene.2012.10.041.
[2] Salmasi FR, Hosseinzadeh M. Power management of an isolated hybrid AC/DC
micro-grid with fuzzy control of battery banks. IET Renew Power Gener 2015;9:484–93.
https://doi.org/10.1049/iet-rpg.2014.0271.
[3] Justo JJ, Mwasilu F, Lee J, Jung J-W. AC-microgrids versus DC-microgrids with
distributed energy resources: a review. Renew Sustain Energy Rev 2013;24:387–405.
https://doi.org/10.1016/J.RSER.2013.03.067.
[4] Arul PG, Ramachandaramurthy VK, Rajkumar RK. Control strategies for a hybrid
renewable energy system: a review. Renew Sustain Energy Rev 2015;42:597–608.
https://doi.org/10.1016/j.rser.2014.10.062.
[5] Hybrid AC/DC Microgrids: A Bridge to Future Energy Distribution Systems –
RWTH AACHEN University Institute for Automation of Complex Power Systems –
English, http://www.acs.eonerc.rwth-aachen.de/cms/E-ON-ERC-ACS/Forschung/
AbgeschlosseneProjekte/~euwe/HYBRID-AC-DC-MICROGRIDS-A-BRIDGE-
TOFUT/?lidx=1 [accessed October 26, 2017].
[6] http://www.electricalindia.in/blog/post/id/8813/a-necessity-of-hybrid-acdcmicrogrids-
inindian-electricity-sector [accessed October 28, 2017].
[7] Ren21, Renewables. “Global status report.” Renewable energy policy network for the
21st century. http://www.ren21.net [accessed June 1, 2018].
[8] P. P. Biswas, P. N. Suganthan, and G. A. J. Amaratunga, ‘‘Optimal power flow
solutions incorporating stochastic wind and solar power,’’ Energy Convers. Manage., vol.
148, pp. 1194–1207, Sep. 2017.
[9] S. S. Reddy, ‘‘Multi-objective optimal power flow for a thermal-wind-solar power
system,’’ J. Green Eng., vol. 7, no. 4, pp. 451–476, Apr. 2018.
[10] K. Zehar and S. Sayah, ‘‘Optimal power flow with environmental constraint using a
fast successive linear programming algorithm: Application to the Algerian power
system,’’ Energy Convers. Manage., vol. 49, no. 11, pp. 3362–3366, Nov. 2008.
[11] Antonio, G.E. (2003) Sistemas Eléctricos de Potencia. Pearson Education, Madrid.
[12] Fernandez, G., Cesar, J., Lopez, F. and Smith, E. (2011) Optimal Load Flow Model
Using Optimization Techniques. Central American University . “José Simeon Cañas”.

58
[13] Lee, Y.-D., Chang, Y.-R., Chan, C.-M. and Ho, Y.-H. (2012) Preliminary
Implementation of Microgrid with Photovoltaic and Microturbine of Stand Alone
Operation. Industry Applications Society Annual Meeting, Las Vegas, 7-11 October
2012, 1-9.
[14] Levron, Y., Guerero, J.M. and Beck, Y. (2013) Optimal Power Flow in Microgrids
with Energy Storage. IEEE F. A. Heredia et al. 458 Transactionon on Power System, 28,
3226-3234.
[15] Dall´Anese, E., Zhu, H. and Giannakis, G.B. (2013) Distributed Optimal Power
Flow for Smart Microgrids. IEEE Transactions on Smart Grid, 4.
[16] Pu, W. and Lemmon, M.D. (2010) Optimal Power Flow in Microgrids Using
Evente-Triggered Optimization. American Control Conference (ACC), Baltimore, 30
June-2 July 2010, 2521-2526. [17] Roa-Sepulveda, B.J. and Pavez-Lazo, C.A. (2001) A
Solution to the Optimal Flow Using Simulated Annelating. IEEE Porto Power Tech
Conference.
[18] Sortomme, E. and El-Sharkawi, M.A. (2009) Optimal Power Flow for a System of
Microgrids. Power Systems Conference and Exposition, Seattle, 15-18 March 2009, 1-5.
[19] A. F. Attia, R. A. El Sehiemy, and H. M. Hasanien, ‘‘Optimal power flow solution
in power systems using a novel sine-cosine algorithm,’’ Int. J. Electr. Power Energy
Syst., vol. 99, pp. 331–343, Jun. 2018.
[20] A. E. Chaib, H. R. E. H. Bouchekara, R. Mehasni, and M. A. Abido, ‘‘Optimal
power flow with emission and non-smooth cost functions using backtracking search
optimization algorithm,’’ Int. J. Elect. Power Energy Syst., vol. 81, pp. 64–77, Oct. 2016.
[21] S. Frank and S. Rebennack, ‘‘An introduction to optimal power flow: Theory,
formulation, and examples,’’ IIE Trans., vol. 48, no. 12, pp. 1172–1197, Aug. 2016.
[22] P. P. Biswas, P. N. Suganthan, and G. A. J. Amaratunga, ‘‘Optimal power flow
solutions incorporating stochastic wind and solar power,’’ Energy Convers. Manage., vol.
148, pp. 1194–1207, Sep. 2017.
[23] S. S. Reddy, ‘‘Multi-objective optimal power flow for a thermal-wind-solar power
system,’’ J. Green Eng., vol. 7, no. 4, pp. 451–476, Apr. 2018.
[24] A. M. Shaheen, R. A. El-Sehiemy, and S. M. Farrag, ‘‘Solving multiobjective
optimal power flow problem via forced initialised differential evolution algorithm,’’ IET
Gener., Transmiss. Distrib., vol. 10, no. 7, pp. 1634–1647, May 2016.
[25] A. R. Bhowmik and A. K. Chakraborty, ‘‘Solution of optimal power flow using non
dominated sorting multi objective opposition based gravitational search algorithm,’’ Int.
J. Elect. Power Energy Syst., vol. 64, pp. 1237–1250, Jan. 2015.

59
[26] K. Zehar and S. Sayah, ‘‘Optimal power flow with environmental constraint using a
fast successive linear programming algorithm: Application to the Algerian power
system,’’ Energy Convers. Manage., vol. 49, no. 11, pp. 3362–3366, Nov. 2008.
[27] P. Fortenbacher and T. Demiray, ‘‘Linear/quadratic programming-based optimal
power flow using linear power flow and absolute loss approximations,’’ Int. J. Electr.
Power Energy Syst., vol. 107, pp. 680–689, May 2019.
[28] S. Mhanna and P. Mancarella, ‘‘An exact sequential linear programming algorithm
for the optimal power flow problem,’’ IEEE Trans. Power Syst., early access, Jul. 14,
2021, doi: 10.1109/TPWRS.2021.3097066.
[29] H. Ambriz-Perez, E. Acha, and C. R. Fuerte-Esquivel, ‘‘Advanced SVC models for
Newton-Raphson load flow and Newton optimal power flow studies,’’ IEEE Trans.
Power Syst., vol. 15, no. 1, pp. 129–136, Feb. 2000.
[30] B. Liu, J. Li, H. Ma, and Y. Liu, ‘‘Generalized benders decomposition based
dynamic optimal power flow considering discrete and continuous decision variables,’’
IEEE Access, vol. 8, pp. 194260–194268, 2020.
[31] R. S. Salgado and E. L. Rangel, ‘‘Optimal power flow solutions through multi-
objective programming,’’ Energy, vol. 42, no. 1, pp. 35–45, Jun. 2012.
[32] V. Vasylius, A. Jonaitis, S. Gudžius, and V. Kopustinskas, ‘‘Multi-period optimal
power flow for identification of critical elements in a country scale high voltage power
grid,’’ Rel. Eng. Syst. Saf., vol. 216, Dec. 2021, Art. no. 107959.
[33] S. Tu, A. Wachter, and E. Wei, ‘‘A two-stage decomposition approach for AC
optimal power flow,’’ IEEE Trans. Power Syst., vol. 36, no. 1, pp. 303–312, Jan. 2021.
[34] M. Pourakbari-Kasmaei, M. Lehtonen, M. Fotuhi-Firuzabad, M. Marzband, and J. R.
S. Mantovani, ‘‘Optimal power flow problem considering multiple-fuel options and
disjoint operating zones: A solverfriendly MINLP model,’’ Int. J. Electr. Power Energy
Syst., vol. 113, pp. 45–55, Dec. 2019.
[35] F. Capitanescu and L. Wehenkel, ‘‘Experiments with the interior-point method for
solving large scale optimal power flow problems,’’ Electr. Power Syst. Res., vol. 95, pp.
276–283, Feb. 2013.
[36] H. Wei, H. Sasaki, J. Kubokawa, and R. Yokoyama, ‘‘An interior point nonlinear
programming for optimal power flow problems with a novel data structure,’’ IEEE Trans.
Power Syst., vol. 13, no. 3, pp. 870–877, Aug. 1998.
[37] E. P. de Carvalho, A. dos Santos, and T. F. Ma, ‘‘Reduced gradient method
combined with augmented Lagrangian and barrier for the optimal power flow problem,’’
Appl. Math. Comput., vol. 200, no. 2, pp. 529–536, Jul. 2008.

60
[38] X. Yuan, B. Zhang, P. Wang, J. Liang, Y. Yuan, Y. Huang, and X. Lei, ‘‘Multi-
objective optimal power flow based on improved strength Pareto evolutionary
algorithm,’’ Energy, vol. 122, pp. 70–82, Mar. 2017.
[39] N. Daryani, M. T. Hagh, and S. Teimourzadeh, ‘‘Adaptive group search
optimization algorithm for multi-objective optimal power flow problem,’’ Appl. Soft
Comput., vol. 38, pp. 1012–1024, Jan. 2016.
[40] K. Abaci and V. Yamacli, ‘‘Differential search algorithm for solving multiobjective
optimal power flow problem,’’ Int. J. Electr. Power Energy Syst., vol. 79, pp. 1–10, Jul.
2016.
[41] H. R. E. H. Bouchekara, A. E. Chaib, M. A. Abido, and R. A. El-Sehiemy, ‘‘Optimal
power flow using an improved colliding bodies optimization algorithm,’’ Appl. Soft
Comput., vol. 42, pp. 119–131, May 2016.
[42] A.-A. A. Mohamed, Y. S. Mohamed, A. A. M. El-Gaafary, and A. M. Hemeida,
‘‘Optimal power flow using moth swarm algorithm,’’ Electr. Power Syst. Res., vol. 142,
pp. 190–206, Jan. 2017.
[43] H. Pulluri, R. Naresh, and V. Sharma, ‘‘Application of stud krill herd algorithm for
solution of optimal power flow problems,’’ Int. Trans. Electr. Energy Syst., vol. 27, no. 6,
p. e2316, Jun. 2017.
[44] H. Pulluri, R. Naresh, and V. Sharma, ‘‘A solution network based on stud krill herd
algorithm for optimal power flow problems,’’ Soft Comput., vol. 22, no. 1, pp. 159–176,
Jan. 2018.
[45] P. P. Biswas, P. N. Suganthan, R. Mallipeddi, and G. A. J. Amaratunga, ‘‘Optimal
power flow solutions using differential evolution algorithm integrated with effective
constraint handling techniques,’’ Eng. Appl. Artif. Intell., vol. 68, pp. 81–100, Feb. 2018.
[46] A. Gacem and D. Benattous, ‘‘Hybrid genetic algorithm and particle swarm for
optimal power flow with non-smooth fuel cost functions,’’ Int. J. Syst. Assurance Eng.
Manage., vol. 8, no. S1, pp. 146–153, Jan. 2017.
[47] H. R. E. H. Bouchekara, A. E. Chaib, and M. A. Abido, ‘‘Optimal power flow using
GA with a new multi-parent crossover considering: Prohibited zones, valve-point effect,
multi-fuels and emission,’’ Electr. Eng., vol. 100, no. 1, pp. 151–165, Mar. 2018.
[48] T. T. Nguyen, ‘‘A high performance social spider optimization algorithm for optimal
power flow solution with single objective optimization,’’ Energy, vol. 171, pp. 218–240,
Mar. 2019.
[49] M. A. Taher, S. Kamel, F. Jurado, and M. Ebeed, ‘‘Modified grasshopper
optimization framework for optimal power flow solution,’’ Elect. Eng., vol. 101, no. 1,
pp. 121–148, Apr. 2019.

61
[50] M. A. Taher, S. Kamel, F. Jurado, and M. Ebeed, ‘‘An improved mothflame
optimization algorithm for solving optimal power flow problem,’’ Int. Trans. Elect.
Energy Syst., vol. 29, no. 3, p. e2743, Mar. 2019.
[51] P. P. Biswas, P. N. Suganthan, R. Mallipeddi, and G. A. J. Amaratunga, ‘‘Multi-
objective optimal power flow solutions using a constraint handling technique of
evolutionary algorithms,’’ Soft Comput., vol. 24, no. 4, pp. 2999–3023, Feb. 2020.
[52] G. Chen, J. Qian, Z. Zhang, and S. Li, ‘‘Application of modified pigeoninspired
optimization algorithm and constraint-objective sorting rule on multi-objective optimal
power flow problem,’’ Appl. Soft Comput., vol. 92, Jul. 2020, Art. no. 106321.
[53] H. Buch and I. N. Trivedi, ‘‘An efficient adaptive moth flame optimization
algorithm for solving large-scale optimal power flow problem with POZ, multifuel and
valve-point loading effect,’’ Iranian J. Sci. Technol., Trans. Electr. Eng., vol. 43, no. 4,
pp. 1031–1051, Dec. 2019.
[54] R. Roy and H. T. Jadhov, ‘‘Optimal power flow solution of power system
incorporating stochastic wind power using Gbest guided artificial bee colony algorithm,’’
Int. J. Elect. Power Energy Syst., vol. 64, pp. 562–578, Jan. 2015.
[55] A. Panda and M. Tripathy, ‘‘Optimal power flow solution of wind integrated power
system using modified bacteria foraging algorithm,’’ Int. J. Elect. Power Energy Syst.,
vol. 54, pp. 306–314, Jan. 2014.
[56] A. Panda and M. Tripathy, ‘‘Security constrained optimal power flow solution of
wind-thermal generation system using modified bacteria foraging algorithm,’’ Energy,
vol. 93, pp. 816–827, Dec. 2015.
[57] S. R. Salkuti, ‘‘Optimal power flow using multi-objective glowworm swarm
optimization algorithm in a wind energy integrated power system,’’ Int. J. Green Energy,
vol. 16, no. 15, pp. 1547–1561, Dec. 2019.
[58] L. Shi, C. Wang, L. Yao, Y. Ni, and M. Bazargan, ‘‘Optimal power flow solution
incorporating wind power,’’ IEEE Syst. J., vol. 6, no. 2, pp. 233–241, Jun. 2012.
[59] I. U. Khan, N. Javaid, K. A. A. Gamage, C. J. Taylor, S. Baig, and X. Ma,
‘‘Heuristic algorithm based optimal power flow model incorporating stochastic
renewable energy sources,’’ IEEE Access, vol. 8, pp. 148622–148643, 2020.
[60] M. A. Ilyas, G. Abbas, T. Alquthami, M. Awais, and M. B. Rasheed, ‘‘Multi-
objective optimal power flow with integration of renewable energy sources using fuzzy
membership function,’’ IEEE Access, vol. 8, pp. 143185–143200, 2020.
[61] S. Li, W. Gong, L. Wang, X. Yan, and C. Hu, ‘‘Optimal power flow by means of
improved adaptive differential evolution,’’ Energy, vol. 198, May 2020, Art. no. 117314.

62
[62] J. Ben Hmida, T. Chambers, and J. Lee, ‘‘Solving constrained optimal power flow
with renewables using hybrid modified imperialist competitive algorithm and sequential
quadratic programming,’’ Electr. Power Syst. Res., vol. 177, Dec. 2019, Art. no. 105989.
[63] E. E. Elattar and S. ElSayed, ‘‘Modified JAYA algorithm for optimal power flow
incorporating renewable energy sources considering the cost, emission, power loss and
voltage profile improvement,’’ Energy, vol. 178, pp. 598–609, Jul. 2019.
[64] Z. Ullah, S. Wang, J. Radosavljević, and J. Lai, ‘‘A solution to the optimal power
flow problem considering WT and PV generation,’’ IEEE Access, vol. 7, pp. 46763–
46772, 2019.
[65] M. H. Sulaiman and Z. Mustaffa, ‘‘Optimal power flow incorporating stochastic
wind and solar generation by metaheuristic optimizers,’’ Microsyst. Technol., vol. 27, pp.
3263–3277, Oct. 2020.
[66] T. Samakpong, W. Ongsakul, and N. Madhu Manjiparambil, ‘‘Optimal power flow
incorporating renewable uncertainty related opportunity costs,’’ Comput. Intell., to be
published.
[67] S. Duman, S. Rivera, J. Li, and L. Wu, ‘‘Optimal power flow of power systems with
controllable wind-photovoltaic energy systems via differential evolutionary particle
swarm optimization,’’ Int. Trans. Elect. Energy Syst., vol. 30, no. 4, Apr. 2020, Art. no.
e12270.
[68] A. Panda, U. Mishra, M.-L. Tseng, and M. H. Ali, ‘‘Hybrid power systems with
emission minimization: Multi-objective optimal operation,’’ J. Cleaner Prod., vol. 268,
Sep. 2020, Art. no. 121418.
[69] M. A. M. Shaheen, H. M. Hasanien, S. F. Mekhamer, and H. E. A. Talaat, ‘‘Optimal
power flow of power systems including distributed generation units using sunflower
optimization algorithm,’’ IEEE Access, vol. 7, pp. 109289–109300, 2019.

63