VOLUME 21(4), 2022 403 Date of publication DEC-31, 2022, date of current version NOV-15, 2022. www.computingonline.net / computing@computingonline.net Print ISSN 1727-6209 Online ISSN 2312-5381 DOI 10.47839/ijc.21.4.2775 Optimum Reactive Power Dispatch Solution using Hybrid Particle Swarm Optimization and Pathfinder Algorithm SAMSON A. ADEGOKE, YANXIA SUN Department of Electrical and Electronic Engineering Science, University of Johannesburg, South Africa, e-mail (adsam4u@gmail.com, ysun@uj.ac.za) Corresponding author: Samson A. Adegoke (e-mail: adsam4u@gmail.com). South Africa National Research Foundation partially supports this research work with the grant (no. 112108 and 112142), and reward grant from South African National Research Foundation with the grant (no. 95687 and 114911), Eskom Tertiary Education Support Programme grant, research grants from URC of University of Johannesburg, and grant of Global Excellence and Stature (GES) University of Johannesburg, South Africa. ABSTRACT Optimum reactive power dispatch (ORPD) significantly impacts the operation and control of electrical power systems (EPS) due to its undeniable benefit in the economic operation and reliability of the systems. ORPD is a sub-problem of optimal power flow (OPF). The main aim is to reduce/minimize the real power loss. Among the swarm intelligence (SI) metaheuristic algorithms is particle swarm optimization (PSO), which has fast convergence speed and gives the optimum solution to a particular problem by moving the swarm in the intensification (exploitation) search space. Also, the pathfinder algorithm (PFA) mimics the collective movement of the swarms with a leading member. Therefore, combining the fast convergence of PSO with PFA to form a hybrid technique is considered a viable approach in this study to avoid decreasing PFA searchability when the dimension of the problem increases. In this article, a hybrid algorithm based on a particle swarm optimization and pathfinder algorithm (HPSO-PFA) is proposed for the first time to study the combination of the control variables (generator voltage, transformer tap, and sizing of reactive compensation to obtain the objective function (total real power loss). The proposed method is tested on the IEEE 30 and 118 bus systems. The losses were reduced to 16.14262 MW and 107.2913 MW for the IEEE 30 and 118 test systems. Furthermore, the percentage (%) reduction for the IEEE 30 and 118 test systems are 9.8% and 19.25%, respectively. The result demonstrates the performance of HPSO-PFA gives a better solution than the other algorithms. KEYWORDS optimum reactive power dispatch HPSO-PFA; pathfinder algorithm (PFA); minimization of power loss. I. INTRODUCTION UE to modern equipment running on electricity, electrical power systems (EPS) have undergone several disturbances. As the demand for electricity goes higher, consumption will gradually be higher. EPS is a process of generating, transmitting, and distributing electric energy to the consumers (house, industry, and transportation use). Indisputably, optimum reactive power dispatch ORPD significantly impacts the operation and control of EPS due to its undeniable benefit in economic operation, security, and reliability of the systems. ORPD is a sub-problem of OPF; it is a nonlinear optimization problem in a power system involving continuous and discrete control variables while obeying the equality and inequality constraints [1–6]. The change in reactive power generation (RPG) on every load variation in power system operation leads to varying/changes in load voltage. However, proper/adequate reactive power management will maintain the voltage profile at each bus/node. The main objective of ORPD is the reduction/minimizing of actual/real power loss while keeping the power balance equality and inequality constraints. The control variables in achieving the objective function are the transformer tap settings, generator voltage magnitude, and shunt capacitors. Moreover, improvement in voltage profile leads to a reduction in real power loss [6]. Many methods have been reported in the literature in finding the solution to the ORPD problem; such techniques include conventional/traditional methods and meta-heuristic methods. Some of the conventional methods are gradient- based, Newton methods, interior point method [7], nonlinear programming, quadratic programming [8], and linear programming [9, 10]. However, these methods are not accurate in dealing with discrete variables and nonlinear functions [1, 11]. Among the meta-heuristic methods which has high quality solutions are: particle swarm optimization (PSO) [12], tight- D