International Journal of Applied Power Engineering (IJAPE)

Vol. 13, No. 4, December 2024, pp. 963~972

ISSN: 2252-8792, DOI: 10.11591/ijape.v13.i4.pp963-972  963

Weighted sum method based multi-objective optimal power

flow considering various objectives: an application of whale
optimization algorithm

Tentu Papi Naidu1,2, Ganapathy Balasubramanian3, Venkateswara Rao Bathina4

1
Lendi Institute of Engineering and Technology, Vijayanagaram, India
2
Department of Electrical Engineering, Faculty of Engineering and Technology, Annamalai University, Chidambaram, India
3
Department of Electrical Engineering, Government College of Engineering, Tirunelveli, India
4
Department of Electrical and Electronics Engineering, V. R. Siddhartha Engineering College,
Deemed to be University, Vijayawada, India

Article Info ABSTRACT

Article history: Nowadays, multi-objective optimization plays a vital role in solving optimal
power flow problems. Multi-objective optimal power flow (MOOPF) is a
Received Dec 31, 2023 nonlinear optimization problem aimed at optimizing control variables while
Revised Aug 7, 2024 balancing multiple objective functions and satisfying both equality and
Accepted Aug 15, 2024 inequality constraints and addresses this by integrating two more objectives
into a single objective using a weighting factor. In this paper this weighted
sum type multi-objective technique has been used to formulate the objective
Keywords: function. The whale optimization algorithm (WOA) has been used to reduce
the cost, emission, losses, and voltage stability by considering various multi
Optimal power flow objectives like fuel cost along with emission, fuel cost with losses, fuel cost
Real power losses with voltage stability, fuel cost with voltage deviation and finally fuel cost
Voltage deviation with emission, losses, voltage deviation. In this paper, the IEEE 30 bus
Voltage stability structure has been used to analyze the effect of WOA on the improvement of
Whale optimization algorithm system performance. Obtained results with WOA have been compared with
other optimization techniques like ensemble constraint handling technique
with differential evolution (ECHT-DE), the superiority of feasible differential
evolution (SF-DE), moth swarm algorithm (MSA), and moth-flame
optimization (MFO), available in the literature.
This is an open access article under the CC BY-SA license.

Corresponding Author:
Venkateswara Rao Bathina
Department of Electrical and Electronics Engineering, V. R. Siddhartha Engineering College
Deemed to be University
Vijayawada, Andhra Pradesh, India
Email: [email protected]

1. INTRODUCTION
In the power system deregulation market, the optimal power flow (OPF) problem is very crucial. This
issue is non-linear, static, controllable, large-scale and convex, non-convex type that optimizes based on
objective function and its solving efficiency with limitations imposed on the power system model, lines, busses
and all equipment’s to satisfy all operating and physical constraints. There will be equality and inequality
constraints to balance all the nodal power flow representations and limitations to control all the state variables
involved. These variables are generator active and reactive powers, its bus voltages, transformer tap changing
are considered as controllable parameters. The load reactive powers, load bus voltages, real and reactive power
flow in the transmission lines are considered as load busses. The economic operation, optimal sharing of power

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

964  ISSN: 2252-8792

between the sources and to the loads meeting all the constraints and also to meet the electric utilities and firms
needs most optimally is referred to as OPF [1].
In the last few years, various bio-inspired optimization OPF algorithms are proposed by many authors
to solve very effectively and easily large complex and multi-objective (MO), multi-constrained problems [2].
The trial and error methods are involved in solving these OPF to achieve the tolerance based optimal
solution(s). The population or bio-inspired optimization problems developed and found to give most optimal
desired solutions [3]. The bio-inspired algorithms are classified as four classes namely, evolution based, swarm
intelligence [4], ecology and multi-objective based. The evolutionary OPF problems are artificial neural
networks [5], genetic algorithms, evolution strategies [6], differential evolution and paddy-field algorithm [7].
The particle swarm, ant-colony [8], artificial bee, fish swarm, bacterial forging [9], fire-fly [10], group-search,
artificial immune system [11], shuffled frog-leaping are famous methods in multi-objective OPF swarm
optimization algorithms. In the ecology based OPF algorithms, invasive weed [12], bio-geography, multiple-
species co-evolution [13] are few important types. The more advanced OPF methods are multi-objective bio-
inspired algorithms such as nondominated sorting genetic algorithm (NSGA-II) method [14], population based
ant-colony [15], strength-pareto, vector evaluated GA, pareto archived evolutionary strategy algorithms [16].
The differential evaluation, solved based on minimizing fuel-cost, increasing voltage stability and
voltage profile. Modified differential evolution [17] algorithm is a non-smooth and non-convex technique for
optimal fuel-cost constraints for a large power system network. An improved scatter search [18] technique is
used to solve environmental and economic power dispatch problem to solve large network with multiple
objectives and constraints. Pareto dominance and crowding distance based neo control method [19], enhanced
genetic algorithm [20], decoupled quadratic load flow [21] for solving optimally fuel cost, line losses and
voltage stability index. A distributed and parallel OPF algorithm for effective use of renewable energy sources
(RES) in smart grid network with fuel cost minimization and carbon emission reduction as constraints to solve
OPF problem. The biogeography-based optimization based on heuristic optimization algorithm to solve
convex/non-convex fuel cost characteristics for OPF problem [22]. Modified shuffle frog leaping algorithm to
solve emission & financial issues and fuzzy evolutionary and particle swarm optimization hybrid scheme for
getting solution to OPF problem with fuel expenditure with various non-linear and linear constraints. Multi-
objective harmony search technique, fast nondominated sorting GA (NSGA-II) technique [23], artificial bee
colony algorithm [24] with multiple linear and non-linear, balanced and unbalanced constraints with multiple
objectives to solve convex and non-convex fuel-price minimizing, environment-friendly with lowering carbon
and other flue-gasses emission, voltage profile and stability enhancement, real power loss decreasing, and
reactive power optimizing as major constraints. Firefly [25] is a hybrid new and effective algorithm, that
improved particle swam optimization (PSO) for multi-objective OPF (MOOPF) issue considering the cost,
voltage stability index, emission, and power loss [26].
The fuzzy adaptive chaotic ant swarm hybrid optimization with sequential quadratic programming
technique employed for resolving economic load dispatch (ELD) issues. Gravitational search method with
various objective functions for the minimization of fuel price, stability of the voltage and enhancement of
profile [27]. The neo hybrid optimization technique employed for modified PSO and shuffled frog leaping
algorithm (SFLA) called as MPSO-SFLA obtain OPF solution under the limitations like forbidden zones and
valve point effect demonstrate their technique is effective in obtaining solution for OPF and ELD problem in
the power systems. This method is found to be effective in improving the overall system profile meeting all
the constraints compared to the earlier methods.
In this paper, five major objective functions like fuel cost, emission, true power losses and voltage
stability and voltage deviation of the network are taken attention in planning of power system that is employed
in whale optimization algorithm. This method is very strong, effective with superior speed to attain the outputs
compared to earlier techniques. Also, with increase in the network size and constraints, its effectiveness also
increases as compared with earlier methods. This is because, the method is a group algorithm and other reason
is because of colonial groups competition based algorithm. The whale algorithm technique is estimated on the
standard Institute of Electrical and Electronics Engineers thirty bus system. The work is studied under different
combinations of five objectives and the best compromise solution is detailed here. The multi-objective OPF
issues shows suggested whale technique is best while comparing to earlier techniques. This paper is
categorized as five sections: section 2 involves in a multi-objective issues formulation section 3 demonstrates
about architecture of whale optimization technique, section 4 is allocated for the results and performance
analysis mentioned methods which are employed to encounter the literature studies of multi-objective OPF
problem on IEEE thirty bus system and finally, in section 5, the conclusion of the implementation for the
proposed technique is presented

Int J Appl Power Eng, Vol. 13, No. 4, December 2024: 963-972
Int J Appl Power Eng ISSN: 2252-8792  965

2. MATHEMATICAL FORMULATION OF MULTI OBJECTIVE OPTIMAL POWER FLOW

(MOOPF) PROBLEMS
Multi-objective optimal power flow (MOOPF) is nonlinear optimization issue . the primary focus is
to optimize control variables while addressing two or more objective functions, while also satisfying both
equality and inequality constraints. This paper accomplishes the integration of two objectives converts in one
objective by introducing a weighting factor as crucial consideration.

2.1. Objective 1: cost minimization

The sum of cost function for fuel is set of generating units is represented in the following equation.
The initial objective function aims to minimize the generation cost [28]. In (1), α, β, and γ are the cost
coefficients of thermal power plants.

𝐹1 = (∑𝑁𝑇𝐺 2
𝑖=1 𝛼𝑖 + 𝛽𝑖 𝑃𝑇𝐺𝑖 + 𝛾𝑖 𝑃𝑇𝐺𝑖 )$/𝐻𝑟 (1)

2.2. Objective 2: minimization of emission

The warm generator delivers the discharge of SOx, NOx with contaminates the environment. Thus, it
is needed to decrease the emanation by accepting this one as an objective. In (2), 𝑎, 𝑏, 𝑐, 𝑑, and 𝑒 are the
emission coefficients of the thermal generation unit.

𝐹2 = ∑𝑁𝑇𝐺
𝑖=1 10
−2 (𝑎 2
𝑖 + 𝑏𝑖 𝑃𝑇𝐺𝑖 + 𝑐𝑖 𝑃𝑇𝐺𝑖 ) + 𝑑𝑖 𝑒𝑥𝑝(𝑒𝑖 𝑃𝑇𝐺𝑖 ) (2)

2.3. Objective 3: minimization of actual power losses

These are calculated employing the (3) [29]. In (3), 𝑉𝑖 is the voltage at ith bus, 𝑉𝑗 is the voltage at jth
bus. NT is the number of transmission lines.
2 2
𝐹3 = ∑𝑁𝑇
𝑘=1 𝐺𝑘(𝑖,𝑗) [𝑉𝑖 + 𝑉𝑗 − 2𝑉𝑖 𝑉𝑗 𝑐𝑜𝑠( 𝛿𝑖𝑗 ) (3)

2.4. Objective 4: voltage stability

To improve the voltage stability in the system, the L-index is calculated for all heap transports, with
the highest value among them serving as the global indicator for system stability. In this manner, the main
focus of system stability is defined as (4) [30].

𝑉𝑖
𝐹4 = |1 − ∑𝑁𝐺
𝑖=1 𝐹𝑗𝑖 | → 𝑤ℎ𝑒𝑟𝑒 → 𝑗 = 1,2, . . . . , 𝑁𝐿 𝑎𝑛𝑑 → 𝐹𝑗𝑖 = −𝑖𝑛𝑣[𝑌𝐿𝐿 ][𝑌𝐿𝐺 ] (4)
𝑉𝑗

2.5. Objective 5: minimization of voltage deviation

Voltage deviation has been determined using (5), here 𝑉𝑛 is the voltage at node n, and 1 is considered
as reference voltage.

𝐹5 = 𝑉𝐷 = ∑𝑁𝑏
𝑛=1 |𝑉𝑛 − 1| (5)

Considering the previously mentioned mono objectives, multiple objectives have been obtained in several
technical studies [13].

2.6. Case 1: reduction of fuel cost and emission

The formulation of the objective function, containing of fuel cost and emission, and the selected
weight factor is 100. In (6), F1 is the objective1 which is reduction cost and F2 is the objective2 which is
emission reduction. These are combined with weighting factor W1.

𝐹𝐹1(𝑋, 𝑈) = 𝐹1 + 𝑊1 ∗ 𝐹2 (6)

2.7. Case 2: curtailment of fuel cost and losses

In the power systems operation transmission loss is the most important element to determine
effectiveness. To minimize the transmission loss in the network together with minimization of cost generation.
The formulation of the objective function, which consists of fuel costs and losses with a chosen weight factor,
is 40.

𝐹𝐹2(𝑋, 𝑈) = 𝐹1 + 𝑊2 ∗ 𝐹3 (7)

Weighted sum method based multi-objective optimal power flow considering … (Tentu Papi Naidu)
966  ISSN: 2252-8792

2.8. Case 3: reduction of fuel cost and improvement of voltage stability

This objective function is focused to reduce cost of fuel while improving system voltage stability. The
multiple objectives are consolidated into one objectively as (8).

𝐹𝐹3(𝑋, 𝑈) = 𝐹1 + 𝑊3 ∗ 𝐹4 (8)

Taken the weight factor is hundred from [9].

2.9. Case 4: reduction of cost of fuel and voltage deviation

The main focus of the objective function is to reduce system's voltage variation and fuel expense.
Multiple objective functions are reduced to a single goal as (9).

𝐹𝐹4(𝑋, 𝑈) = 𝐹1 + 𝑊4 ∗ 𝐹5 (9)

The weight factor 100 is taken [10].

2.10. Case 5: minimization of fuel cost, emission, voltage deviation and losses
This case study combines four objective functions. The simultaneous minimization of fuel cost,
emissions, voltage variation, and real power loss in the network. The objective function is given by (10).

𝐹𝐹5(𝑋, 𝑈) = 𝐹1 + 𝑊5 ∗ 𝐹2 + 𝑊6 ∗ 𝐹3 + 𝑊7 ∗ 𝐹5 (10)

W5=19, W6=21, and W7=22 are taken to balance between the objectives.

2.11. Equality constraints

Basic load flow equations like these require that the power produced match the power demand and
losses [28]. In below equations 𝑃𝐺𝑖 is the true power generation and 𝑃𝐷𝑖 is the true power demand. 𝑄𝐺𝑖 is the
imaginary power generation and 𝑄𝐷𝑖 is the imaginary power demand.

𝑁𝑏 𝐺𝑖𝑗 𝑐𝑜𝑠 𝜃𝑖𝑗

𝑃𝐺𝑖 − 𝑃𝐷𝑖 − 𝑉𝑖 ∑𝑖=1 𝑉𝑗 ( )=0 (11)
+𝐵𝑖𝑗 𝑠𝑖𝑛 𝜃𝑖𝑗

𝑁𝑏 𝐺𝑖𝑗 𝑠𝑖𝑛 𝜃𝑖𝑗

𝑄𝐺𝑖 − 𝑄𝐷𝑖 − 𝑉𝑖 ∑𝑖=1 𝑉𝑗 ( )=0 (12)
𝐵𝑖𝑗 𝑐𝑜𝑠𝜃𝑖𝑗

2.12. Inequality constraints

Maximum and minimum values of generator bus voltages and load bus voltages considered as
inequality restrictions. They get along with imaginary power generation limits, limits of the transformer tap
settings and capacitor banks minimum and maximum values [30].
𝑚𝑖𝑛 𝑚𝑎𝑥
𝑉𝐺𝑖 ≤ 𝑉𝐺𝑖 ≤ 𝑉𝐺𝑖 , 𝑖 ∈ 𝑁𝑔 (13)

𝑉𝐿𝑖𝑚𝑖𝑛 ≤ 𝑉𝐿𝑖 ≤ 𝑉𝐿𝑖𝑚𝑎𝑥 , 𝑖 ∈ 𝑁𝑙 (14)

𝑚𝑖𝑛 𝑚𝑎𝑥
𝑄𝐺𝑖 ≤ 𝑄𝐺𝑖 ≤ 𝑄𝐺𝑖 , 𝑖 ∈ 𝑁𝑡 (15)

𝑇𝑖𝑚𝑖𝑛 ≤ 𝑇𝑚 ≤ 𝑇𝑖𝑚𝑎𝑥 , 𝑖 ∈ 𝑁𝑐 (16)

𝑚𝑖𝑛 𝑚𝑎𝑥
𝑄𝑐𝑖 ≤ 𝑄𝑐𝑖 ≤ 𝑄𝑐𝑖 , 𝑖 ∈ 𝑁𝑐 (17)

3. WHALE OPTIMIZATION ALGORITHM (WOA)

Whales are magnificent creatures, with humpback whales standing out due to their remarkable hunting
strategy known as the bubble-net feeding technique. This foraging behavior involves two distinct maneuvers
known as 'upwinding' and 'double loops.' During the former, humpback whales dive approximately 12 meters
deep and then create a twisting pattern of bubbles around their prey as they ascend toward the surface. The
latter maneuver consists of 3 distinct phases: coral circle, lob tail, and capture circle. You can find more detailed
information about this behavior elsewhere. It's important to emphasize that bubble-net feeding is a unique
behavior exclusive to humpback whales. The bubble net method of the whale is shown in Figure 1 [2].

Int J Appl Power Eng, Vol. 13, No. 4, December 2024: 963-972
Int J Appl Power Eng ISSN: 2252-8792  967
→ →→
𝐷 = | 𝐶 . 𝑋∗ (𝑡) − →
𝑋 (𝑡)| (18)
→(𝑡)
𝑋 =→ → →
𝑋 ∗ (𝑡) − 𝐴. 𝐷| (19)
→
𝐴 = 2 →𝑎. →𝑟 − →𝑎. (20)
→
𝑐 = 2. →𝑟. (21)

Whale, denoted as 'r' is expressed as vector comprising of actual values are written as (22) [27].
𝑇
𝑌𝑖 = (𝑌𝑖,1 , 𝑌𝑖,2 , … . 𝑥𝑖,𝑘 ) 𝑠𝑢𝑏𝑗𝑒𝑐𝑡𝑒𝑑 𝑡𝑜 0 < 𝑥𝑖.1 … … < 𝑥𝑖,𝑘 < 𝐿 (22)

The positions of the whales are determined arbitrarily using (23).

𝑌𝑖,𝑗 = 𝑔𝑚𝑖𝑛 + 𝑟𝑎𝑛𝑑(0,1) ∗ (𝑔𝑚𝑎𝑥 − 𝑔𝑚𝑖𝑛 ) (23)

Control parameter limits are provided in Table 1 and values of WOA are given in Table 2. Steps to
Implementing the WOA to solve the MOOPF:
- Randomly generate initial positions of whales and set algorithm parameters such as population size,
maximum iterations, and convergence criteria.
- Calculate the fitness function value of each whale based on the MOOPF objective function
- Update the positions of the whales using the encircling prey, bubble-net attacking, and search for prey
mechanisms of WOA.
- Repeat the evaluation and update steps for a set number of iterations.
- After convergence, take the values of optimal generator settings, power losses, fuel cost, and fitness
function value.

Figure 1. Bubble net method of whale

Table 1. Control parameters limits

Control parameters Min – Max (p.u)
Generator voltages 0.95 – 1.10
Transformers tap settings 0.90 – 1.10
Shunt capacitors 0.00– 0.20

Table 2. Control - parameters values for WOA

S.No Parameter
Search Agents_no 30
Max_iteration 500
a Linearly is reduced from 02 to 00
r1 and r2 Random numbers in [0,1]

4. RESULTS AND DISCUSSION

In this section, it is explained the results of research and at the same time is given
the comprehensive discussion. Results can be presented in figures, graphs, tables and others that make
the reader understand easily [14], [15]. The discussion can be made in several sub-sections.

Weighted sum method based multi-objective optimal power flow considering … (Tentu Papi Naidu)
968  ISSN: 2252-8792

This paper introduces an optimization approach aimed at minimizing the total cost of real power
generation while considering factors such as losses, emissions, and voltage stability. The proposed method
involves the control of generator bus voltages, the adjustment of reactive power compensation device ratings,
and the optimization of transformer tap settings. Table 3 gives information about IEEE 30 bus system. The cost
coefficient values of generators, bus data, load data, and line data are taken from [27].
From Table 4 it is observed that by optimizing only fuel cost, cost has been reduced to 800.3196 $/hr
but emission is 0.5437 p.u, with emission optimization fuel cost is 944.921 $/hr but emission getting reduced
to 0.2048. By applying single objective optimization that particular objective value became lower but other
objectives have high values, therefore to avoid this multi-objective optimization has been used. From this table
it is also observed that by combining fuel cost and emission provides the moderate values. Here cost is 802.172
$/hr and emission is 0.3293 p.u. From the Table 4 it has been also observed that by combining fuel cost and
losses provides the moderate values. Here cost is 857.81 $/hr and losses are 4.4755 MW. From Table 5 obtained
multi objective values, cost is 800.36 $/hr and voltage stability is 0.1266 p.u. Table 5 also obtained multi
objective values, cost is 800.36 $/hr and deviation of voltage is 0.2011 p.u.

Table 3. The main characteristics of the studied system

IEEE 30
Characteristics Value Details
Buses 30 --
Branches 41 --
Generators 06 Buses: 1, 2, 5, 8, 11, and 13
Load voltage limits 24 [0.94 p.u - 1.06 p.u]
Shunt VAR compensation 09 Buses:10, 12, 15, 17, 20, 21, 23, 24, and 29
Transformers with off-nominal tap ratio 04 Branches: 11, 12, 15, and 36
Control variable 24 --

Table 4. Optimal solutions obtained for combined fuel cost and emission and combined fuel cost and power
losses by WOA for IEEE 30 bus system
Combined fuel cost and emission Combined fuel cost and power losses
In fuel cost In power loss In combined
In emission In combined In fuel cost
minimization minimization (Case 2)
minimization (Case 1) minimization
scheduling scheduling of scheduling of
Control variables and scheduling of scheduling of scheduling of
of generator generator generator
parameters generator units generator units generator units
units and units and units and
and other and other and other
other other other
parameters parameters parameters
parameters parameters parameters
PTG1 176.0386 64.1557 162.75 176.0386 51.299 102.64
PTG2 48.5459 67.6433 51.7207 48.5459 80.0000 54.4114
PTG5 21.2817 50.0000 21.8936 21.2817 50.0000 36.7556
PTG8 21.6116 35.0000 27.1089 21.6116 035 035
PTG11 12.5939 30.0000 13.6142 12.5939 030 29.6401
PTG13 12.1423 40.0000 14.8104 012.1423 040 29.5766
VTG1 01.1 01.10 1.1000 01.1 01.1 0 1.10
VTG2 01.1 01.10 01.10 01.1 01.1 01.10
VTG5 01.1 01.10 01.10 01.1 01.08 01.0838
VTG8 01.08869 01.10 01.0903 01.08869 01.1 01.10
VTG11 01.1 01.10 01.10 01.1 01.1 01.0432
VTG13 01.1 01.10 01.10 01.1 01.1 01.1000
QC10 4.32262 0.4593 0 4.32262 05 3.2100
QC12 0 1.8154 4.6101 0 05 05.0
QC15 0 4.1381 0 0 05 05.0
QC17 2.57489 5.0000 0 2.57489 05 05.0000
QC20 4.11584 5.0000 4.6122 4.11584 05 5.0000
QC21 2.5457 5.0000 1.9727 2.5457 05 5.0000
QC23 1.75619 5.0000 4.6147 1.75619 05 5.0000
QC24 3.97527 5.0000 4.6099 3.97527 05 5.0000
QC29 1.86436 5.0000 4.6108 1.86436 02.5237 5.0000
T11 0.983227 1.1000 1.0022 0.983227 00.9458 0.9740
T12 1.00358 1.1000 1.0022 1.00358 01.10 1.1000
T15 0.992703 1.1000 0.9983 0.992703 00.9960 1.1000
T36 1.00521 1.1000 1.0021 1.00521 00.9849 1.0356
Fuel cost ($/hr) 800.3196 944.921 802.172 800.3196 966.69 857.81
Total power loss (MW) 8.8140 3.399 8.1001 8.8140 2.899 4.4755
Voltage stability p.u 0.1542 0.1455 0.1299 0.1542 0.1260 0.1355
Voltage deviation p.u 1.7624 1.0149 1.6701 1.7624 2.0857 1.3687
Emission p.u 0.5437 0.2048 0.3293 0.5437 0.20724 0.2283
Fitness function value 800.3196 0.2048 834.91 800.3196 2.899 1036.53

Int J Appl Power Eng, Vol. 13, No. 4, December 2024: 963-972
Int J Appl Power Eng ISSN: 2252-8792  969

Table 5. Optimal solutions obtained for combined fuel cost and voltage stability and combined fuel cost and
voltage deviation by WOA for IEEE 30 bus system
Control variables and Combined fuel cost and voltage stability Combined fuel cost and voltage deviation
parameters In fuel cost In voltage In combined In fuel cost In voltage In combined
minimization stability (Case 3) minimization deviation (Case 4)
scheduling of minimization scheduling of scheduling minimization scheduling of
generator scheduling of generator of generator scheduling of generator
units and generator units units and units and generator units units and
other and other other other and other other
parameters parameters parameters parameters parameters parameters
PTG1 176.0386 80.528 175.67 176.0386 127.8870 180.5493
PTG2 48.5459 80.0000 48.0976 48.5459 73.2439 49.0249
PTG5 21.2817 50.0000 20.7299 21.2817 30.8068 23.3954
PTG8 21.6116 35.0000 23.2889 21.6116 15.2479 14.9210
PTG11 12.5939 30.0000 12.2847 12.5939 18.6419 11.8863
PTG13 12.1423 12.0000 12.1561 12.1423 29.5643 14.2723
VTG1 01.1 01.1000 01.1000 1.1 0.9661 1.0298
VTG2 01.1 01.1000 01.0891 01.01 1.0306 1.0106
VTG5 01.1 01.1000 01.0616 01.01 1.0025 1.0019
VTG8 01.08869 01.1000 01.0795 01.08869 1.0288 1.0162
VTG11 01.1 01.1000 01.1000 1.1 1.0596 1.0330
VTG13 01.1 01.1000 01.1000 1.1 1.0098 1.0399
QC10 4.32262 5.0000 0.3549 4.32262 0.2454 0
QC12 0 5.0000 0 0 2.5816 2.0817
QC15 0 5.0000 3.0558 0 1.6725 3.9460
QC17 2.57489 5.0000 0.8241 2.57489 2.2052 1.1293
QC20 4.11584 5.0000 0 4.11584 0 1.9271
QC21 2.5457 5.0000 2.2320 2.5457 4.2276 4.2315
QC23 1.75619 5.0000 2.1422 1.75619 2.7822 0.6253
QC24 3.97527 5.0000 1.7377 3.97527 4.0040 1.1877
QC29 1.86436 5.0000 1.8542 1.86436 4.5123 1.7332
T11 0.983227 0.9000 0.9183 0.983227 0.9560 0.9462
T12 1.00358 0.9000 1.1000 1.00358 1.0546 1.0074
T15 0.992703 0.9000 0.9321 0.992703 0.9474 0.9432
T36 1.00521 0.9000 0.9504 1.00521 0.9713 0.9660
Fuel cost ($/hr) 800.3196 919.692 800.36 800.3196 855.5189 806.105
Total power loss (MW) 8.8140 4.128 8.8442 0.5437 0.2676 0.3751
Voltage stability p.u 0.1542 0.1088 0.1266 8.8140 11.89 10.609
Voltage deviation p.u 1.7624 3.4375 1.7154 1.7624 0.2011 0.2022
Emission p.u 0.5437 0.2250 0.3621 0.1542 0.1463 0.1482
Fitness function value 800.3196 0.1088 813.03 800.3196 0.2011 826.1644

Figure 2 shows the convergence curves for case 1 to case 5. From the Figure 2 it has been observed
that case 5 consisting of multiple objectives produce the compromising solution. Table 6 presents the control
variables of all single objectives and multi objective consisting of all the objectives. From this it is observed
that by combining all the objectives best optimal values have been achieved. Table 7 presents the comparison
of case 2, case 3, and case5 of WOA with other algorithm available in literature. From this table it is observed
that fitness function value with WOA is best compared to ensemble constraint handling technique with
differential evolution (ECHT-DE), superiority of feasible differential evolution (SF-DE), moth swarm
algorithm (MSA), and moth-flame optimization (MFO).

Figure 2. Convergence curves for various cases

Weighted sum method based multi-objective optimal power flow considering … (Tentu Papi Naidu)
970  ISSN: 2252-8792

Table 6. Optimal solutions obtained for combined fuel cost, voltage deviation, power losses and emission
by WOA for IEEE 30 bus system
Control variables and In fuel cost In voltage deviation In power loss In emission In combined
parameters minimization minimization minimization minimization (Case 5)
scheduling of scheduling of scheduling of scheduling of scheduling of
generator units and generator units and generator units generator units generator units
other parameters other parameters and other and other and other
parameters parameters parameters
PTG1 176.0386 127.887 51.299 64.1557 125.49
PTG2 48.5459 73.2439 80.000 67.6433 53.9522
PTG5 21.2817 30.8068 50.000 50.0000 31.0859
PTG8 21.6116 15.2479 35.000 35.0000 35.0000
PTG11 12.5939 18.6419 30.000 30.0000 21.5385
PTG13 12.1423 29.5643 40.000 40.0000 21.7013
VTG1 1.1 0.9661 1.1000 1.1000 1.1000
VTG2 1.1 1.0306 1.1000 1.1000 1.0864
VTG5 1.1 1.0025 1.0862 1.1000 1.0599
VTG8 1.08869 1.0288 1.1000 1.1000 1.0690
VTG11 1.1 1.0596 1.1000 1.1000 1.0832
VTG13 1.1 1.0098 1.1000 1.1000 1.0173
QC10 4.32262 0.2454 5.0000 0.4593 1.6604
QC12 00 2.5816 5.0000 1.8154 4.0284
QC15 00 1.6725 5.0000 4.1381 3.8585
QC17 2.57489 2.2052 5.0000 5.0000 0.1981
QC20 4.11584 00 5.0000 5.0000 4.0745
QC21 2.5457 4.2276 5.0000 5.0000 4.1727
QC23 1.75619 2.7822 5.0000 5.0000 4.2522
QC24 3.97527 4.0040 5.0000 5.0000 4.8502
QC29 1.86436 4.5123 2.5237 5.0000 4.3251
T11 0.983227 0.9560 0.9458 1.1000 1.1000
T12 1.00358 1.0546 1.1000 1.1000 1.0052
T15 0.992703 0.9474 0.9960 1.1000 1.0635
T36 1.00521 0.9713 0.9849 1.1000 1.0478
Fuel cost ($/hr) 800.3196 855.518 966.69 944.921 824.82
Emission p.u 0.5437 0.2676 0.20724 0.2048 0.2584
Total power loss (MW) 8.8140 11.89 2.899 3.399 5.5871
Voltage deviation p.u 1.7624 0.2011 2.0857 1.0149 0.4943
Voltage stability p.u 0.1542 0.1463 0.1260 0.1455 0.1468
Fitness function value 800.3196 0.2011 2.899 0.2048 962.96

Table 7. Comparison of the WOA with ECHT-DE, SF-DE, MSA, and MFO for IEEE 30 bus system
considering various cases
Objective function Objective WOA ECHT-DE SF-DE MSA MFO
Case 5 Fuel cost ($/h) 824.82 830.1156 830.1366 830.639 830.9135
Emission (ton/h) 0.2584 0.25293 0.25313 0.25258 0.25231
PLoss (MW) 5.5871 5.5894 5.5887 5.6219 5.5971
L-index 0.1468 0.14748 0.14756 0.14802 0.14556
Fitness function 962.96 964.1331 964.1254 965.2905 965.8077
Case 3 Fuel cost ($/h) 800.36 800.4321 800.4203 801.2248 801.668
Emission (ton/h) 0.3621 0.36585 0.36592 0.36106 0.34299
PLoss (MW) 8.8442 9.0043 8.9985 8.9761 8.5578
L-index 0.1266 0.13739 0.13745 0.13713 0.13759
Fitness function 813.03 814.1708 814.1649 814.9378 815.4270
Case 2 Fuel cost ($/h) 857.81 858.867 859.1458 859.1915 858.5812
Emission (ton/h) 0.2283 0.22902 0.2289 0.22899 0.22947
PLoss (MW) 4.4755 4.5321 4.5245 4.5404 4.5772
L-index 0.1355 0.13796 0.13785 0.13814 0.13806
Fitness function 1036.53 1040.151 1040.125 1040.808 1041.671

5. CONCLUSION
The whale optimization algorithm (WOA) combined with optimal power flow (OPF) demonstrates
superior performance across multiple objectives, including fuel cost, emissions, losses, voltage stability, and
voltage deviation. From the results, it has been observed that by using weighted sum type multi-objective all
the objectives optimized simultaneously and provided the compromising solution. In case 2, minimizing the
fuel cost along with emission the objective function value is 1036.53 p.u. In the case 3, minimizing the fuel
cost along with losses the objective function value in 813.03 p.u, to get the compromising solution by
combining all the cases the objective function value became 962.96 p.u. it indicated that all the objectives were

Int J Appl Power Eng, Vol. 13, No. 4, December 2024: 963-972
Int J Appl Power Eng ISSN: 2252-8792  971

optimized simultaneously. The results indicate that, when compared to ECHT-DE, SF-DE, MSA, and MFO,
the WOA-based approach with regulated variables consistently delivers superior outcomes. These findings
have been validated using the IEEE 30 bus system. Additionally, future research could explore the integration
of flexible AC transmission system (FACTS) devices to further optimize system performance.

REFERENCES
[1] T. T. Borges, S. Carneiro, P. A. N. Garcia, and J. L. R. Pereira, “A new OPF based distribution system restoration method,”
International Journal of Electrical Power and Energy Systems, vol. 80, pp. 297–305, 2016, doi: 10.1016/j.ijepes.2016.01.024.
[2] S. Mirjalili and A. Lewis, “The Whale Optimization Algorithm,” Advances in Engineering Software, vol. 95, pp. 51–67, 2016, doi:
10.1016/j.advengsoft.2016.01.008.
[3] Y. Li, Y. Li, G. Li, D. Zhao, and C. Chen, “Two-stage multi-objective OPF for AC/DC grids with VSC-HVDC: Incorporating
decisions analysis into optimization process,” Energy, vol. 147, pp. 286–296, 2018, doi: 10.1016/j.energy.2018.01.036.
[4] X. Fan, W. Sayers, S. Zhang, Z. Han, L. Ren, and H. Chizari, “Review and Classification of Bio-inspired Algorithms and Their
Applications,” Journal of Bionic Engineering, vol. 17, no. 3, pp. 611–631, 2020, doi: 10.1007/s42235-020-0049-9.
[5] E. Barocio, J. Regalado, E. Cuevas, F. Uribe, P. Zúñiga, and P. J. R. Torres, “Modified bio-inspired optimisation algorithm with a
centroid decision making approach for solving a multi-objective optimal power flow problem,” IET Generation, Transmission and
Distribution, vol. 11, no. 4, pp. 1012–1022, 2017, doi: 10.1049/iet-gtd.2016.1135.
[6] Y. Muhammad, R. Khan, M. A. Z. Raja, F. Ullah, N. I. Chaudhary, and Y. He, “Design of Fractional Swarm Intelligent Computing
with Entropy Evolution for Optimal Power Flow Problems,” IEEE Access, vol. 8, pp. 111401–111419, 2020, doi:
10.1109/ACCESS.2020.3002714.
[7] X. Pan, T. Zhao, M. Chen, and S. Zhang, “DeepOPF: A deep neural network approach for security-constrained DC optimal power
flow,” IEEE Transactions on Power Systems, vol. 36, no. 3, pp. 1725–1735, 2021, doi: 10.1109/TPWRS.2020.3026379.
[8] S. C. Kim and S. R. Salkut, “Optimal power flow based congestion management using enhanced genetic algorithms,” International
Journal of Electrical and Computer Engineering, vol. 9, no. 2, pp. 875–883, 2019, doi: 10.11591/ijece.v9i2.pp875-883.
[9] A. M. Shaheen, R. A. El-Sehiemy, and S. M. Farrag, “Solving multi-objective optimal power flow problem via forced initialised
differential evolution algorithm,” IET Generation, Transmission and Distribution, vol. 10, no. 7, pp. 1634–1647, 2016, doi:
10.1049/iet-gtd.2015.0892.
[10] G. Guo, J. Qian, and S. Li, “Optimal power flow based on novel multi-objective artificial fish swarm algorithm,” Engineering
Letters, vol. 28, no. 2, pp. 542–550, 2020.
[11] A. Panda, M. Tripathy, A. K. Barisal, and T. Prakash, “A modified bacteria foraging based optimal power flow framework for
Hydro-Thermal-Wind generation system in the presence of STATCOM,” Energy, vol. 124, pp. 720–740, 2017, doi:
10.1016/j.energy.2017.02.090.
[12] M. P. Varghese and A. Amudha, “Enhancing the Efficiency of Wind Power Using Hybrid Fire Fly and Genetic Algorithm -
Economic Load Dispatch Model,” Current Signal Transduction Therapy, vol. 13, no. 1, pp. 3–10, 2018, doi:
10.2174/1574362413666180223125127.
[13] B. V. Rao and G. V. N. Kumar, “Optimal power flow by BAT search algorithm for generation reallocation with unified power flow
controller,” International Journal of Electrical Power and Energy Systems, vol. 68, pp. 81–88, 2015, doi:
10.1016/j.ijepes.2014.12.057.
[14] A. M. Dalavi, P. J. Pawar, and T. P. Singh, “Determination of optimal tool path in drilling operation using modified shuffled frog
leaping algorithm,” International Journal for Engineering Modelling, vol. 32, no. 2–4, pp. 33–44, 2019, doi:
10.31534/engmod.2019.2-4.ri.01v.
[15] Z. X. Zheng, J. Q. Li, and H. Y. Sang, “A hybrid invasive weed optimization algorithm for the economic load dispatch problem in
power systems,” Mathematical Biosciences and Engineering, vol. 16, no. 4, pp. 2775–2794, 2019, doi: 10.3934/mbe.2019138.
[16] S. Gupta, N. Singh, and K. Joshi, “Biogeography based novel AI optimization with SSSC for optimal power flow,” Majlesi Journal
of Electrical Engineering, vol. 12, no. 2, pp. 39–45, 2018.
[17] E. X. S. Araujo, M. C. Cerbantes, and J. R. S. Mantovani, “Optimal Power Flow with Renewable Generation: A Modified NSGA-
II-based Probabilistic Solution Approach,” Journal of Control, Automation and Electrical Systems, vol. 31, no. 4, pp. 979–989,
2020, doi: 10.1007/s40313-020-00596-7.
[18] M. A. A. Rahman, B. Ismail, K. Naidu, and M. K. Rahmat, “Review on population-based metaheuristic search techniques for
optimal power flow,” Indonesian Journal of Electrical Engineering and Computer Science, vol. 15, no. 1, pp. 373–381, 2019, doi:
10.11591/ijeecs.v15.i1.pp373-381.
[19] X. Yuan et al., “Multi-objective optimal power flow based on improved strength Pareto evolutionary algorithm,” Energy, vol. 122,
pp. 70–82, 2017, doi: 10.1016/j.energy.2017.01.071.
[20] W. Warid, H. Hizam, N. Mariun, and N. I. Abdul Wahab, “A novel quasi-oppositional modified Jaya algorithm for multi-objective
optimal power flow solution,” Applied Soft Computing, vol. 65, pp. 360–373, 2018, doi: 10.1016/j.asoc.2018.01.039.
[21] S. Li, W. Gong, C. Hu, X. Yan, L. Wang, and Q. Gu, “Adaptive constraint differential evolution for optimal power flow,” Energy,
vol. 235, p. 121362, Nov. 2021, doi: 10.1016/j.energy.2021.121362.
[22] R. Devarapalli, B. V. Rao, B. Dey, K. V. Kumar, H. Malik, and F. P. G. Marquez, “An approach to solve OPF problems using a
novel hybrid whale and sine cosine optimization algorithm,” Journal of Intelligent and Fuzzy Systems, vol. 42, no. 2, pp. 957–967,
2022, doi: 10.3233/JIFS-189763.
[23] G. Chen, X. Yi, Z. Zhang, and H. Lei, “Solving the multi-objective optimal power flow problem using the multi-objective firefly
algorithm with a constraints-prior pareto-domination approach,” Energies, vol. 11, no. 12, pp. 34–38, 2018, doi:
10.3390/en11123438.
[24] A. Meng et al., “A high-performance crisscross search based grey wolf optimizer for solving optimal power flow problem,” Energy,
vol. 225, p. 120211, Jun. 2021, doi: 10.1016/j.energy.2021.120211.
[25] E. Naderi, M. Pourakbari-Kasmaei, and H. Abdi, “An efficient particle swarm optimization algorithm to solve optimal power flow
problem integrated with FACTS devices,” Applied Soft Computing, vol. 80, pp. 243–262, Jul. 2019, doi:
10.1016/j.asoc.2019.04.012.
[26] L. Bhamidi and S. Shanmugavelu, “Multi-objective Harmony Search Algorithm for Dynamic Optimal Power Flow with Demand
Side Management,” Electric Power Components and Systems, vol. 47, no. 8, pp. 692–702, 2019, doi:
10.1080/15325008.2019.1627599.

Weighted sum method based multi-objective optimal power flow considering … (Tentu Papi Naidu)
972  ISSN: 2252-8792

[27] R. A. El Sehiemy, F. Selim, B. Bentouati, and M. A. Abido, “A novel multi-objective hybrid particle swarm and salp optimization
algorithm for technical-economical-environmental operation in power systems,” Energy, vol. 193, p. 116817, Feb. 2020, doi:
10.1016/j.energy.2019.116817.
[28] M. S. Alkoffash, M. A. Awadallah, M. Alweshah, R. A. Zitar, K. Assaleh, and M. A. Al-Betar, “A Non-convex Economic Load
Dispatch Using Hybrid Salp Swarm Algorithm,” Arabian Journal for Science and Engineering, vol. 46, no. 9, pp. 8721–8740, 2021,
doi: 10.1007/s13369-021-05646-z.
[29] C. Shilaja and T. Arunprasath, “Optimal power flow using Moth Swarm Algorithm with Gravitational Search Algorithm considering
wind power,” Future Generation Computer Systems, vol. 98, pp. 708–715, 2019, doi: 10.1016/j.future.2018.12.046.
[30] E. E. Elattar, “Environmental economic dispatch with heat optimization in the presence of renewable energy based on modified
shuffle frog leaping algorithm,” Energy, vol. 171, pp. 256–269, Mar. 2019, doi: 10.1016/j.energy.2019.01.010.

BIOGRAPHIES OF AUTHORS

Tentu Papi Naidu received his bachelor of engineering degree in electrical and
electronics engineering in 2008. He received his master of engineering degree in power and
industrial drives in 2014 from JNTU Kakinada, A.P, India. He is presently pursuing his Ph.D.
degree in electrical engineering from Annamalai University, Annamalainagar, Tamilnadu.
He is an associate professor in the Electrical and Electronics Engineering Department at
Lendi Institute of Engineering and Technology, Vijayanagaram, AP. His research interests
are in power systems and power electronics. He can be contacted at email:
[email protected].

Ganapathy Balasubramanian received his bachelor of engineering degree in

electrical and electronics engineering in 2000. He received his master of engineering degree
in power system engineering in 2005 from Annamalai University, Annamalainagar,
Tamilnadu, India. He received his Ph.D. degree in electrical engineering from Annamalai
University, Annamalainagar, Tamilnadu, India in 2014. He is associate professor in the
Electrical Engineering Department at Government College of Engineering, Tirunelveli,
Tamilnadu, India. His research interests are in electrical machines, power systems, power
electronics, solar, wind energy, and microgrid applications. He can be contacted at email:
[email protected].

Venkateswara Rao Bathina received his bachelor degree in electrical and

electronics engineering from College of Engineering, Gandhi Institute of Technology and
Management (GITAM), Visakhapatnam, India in 2000, and the master degree in electrical
power engineering from the College of Engineering, JNTU, Hyderabad in 2007 and received
Ph.D. from Jawaharlal Nehru Technological University, Hyderabad in 2015. He is presently
working as an associate professor in the Department of Electrical and Electronics
Engineering, V R Siddhartha Engineering College, Vijayawada. His research interests are
power system stability analysis, FACTS devices, and power system optimization. He has
published several research papers in national and international conferences and journals. He
can be contacted at email: [email protected].

Int J Appl Power Eng, Vol. 13, No. 4, December 2024: 963-972