Optimal Reconfiguration of Radial Distribution Network for Loss Minimization Using Hybrid Genetic Algorithm-Particle Swarm Optimization Algorithm
Oluwagbenga Samuel Alake *
Department of Electrical and Electronic Engineering, Ekiti State University (EKSU), Ado Ekiti, Nigeria.
Emmanuel Taiwo Fasina
Department of Electrical and Electronic Engineering, Ekiti State University (EKSU), Ado Ekiti, Nigeria.
Bankole Adebanji
Department of Electrical and Electronic Engineering, Ekiti State University (EKSU), Ado Ekiti, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
Aim: The aim of this research is to minimize power losses and improve the voltage profile of the Ado-Ekiti 33 kV feeders through optimal network reconfiguration using a Hybrid Genetic Algorithm–Particle Swarm Optimization technique.
Study Design: This study investigates the optimal reconfiguration of the Ado-Ekiti 33 kV radial distribution network using a Hybrid Genetic Algorithm–Particle Swarm Optimization (GA–PSO) approach for technical-loss reduction and voltage-profile improvement. A computational simulation and comparative optimization design was adopted.
Place and Duration of Study: Ado-Ekiti, Nigeria, from August 2025 to July 2026.
Methodology: The Ado-Ekiti network was modelled in MATLAB and evaluated using a Backward/Forward Sweep (BFS) load-flow procedure appropriate for radial distribution feeders. The optimization determined feasible switching configurations subject to bus-voltage, radiality, branch-current, power-balance, connectivity, and switch-status constraints, with a multi-objective fitness function combining active-power-loss and voltage-deviation terms. These constraints ensured that the optimal configuration obtained by the Hybrid GA–PSO algorithm was electrically feasible and operationally practicable. Performance was compared with that of standalone GA and PSO, and the framework was validated using the IEEE 33-bus benchmark.
Results: The Hybrid GA–PSO approach reduced total active power losses by 38.44%, from 2.094 MW to 1.289 MW, while improving the minimum bus voltage from 0.916 p.u. to above 0.95 p.u. The optimization achieved estimated annual energy savings of 3,173.31 MWh, approximately 3.17331 GWh/year, equivalent to approximately ₦131 million in economic benefits, demonstrating improved network efficiency and voltage performance.
Conclusion: The hybridization of GA and PSO effectively leveraged the global-search and diversity-preservation capabilities of GA and the rapid convergence ability of PSO, thereby addressing the individual limitations of each algorithm. The study demonstrates the potential of constrained network reconfiguration using Hybrid GA–PSO to reduce technical losses and improve voltage performance. Total active-power loss decreased from 2.094 MW to 1.289 MW, representing a 38.44% reduction, while the minimum bus voltage improved from 0.916 p.u. to approximately 0.962 p.u.
Keywords: Distribution network reconfiguration, radial distribution network, Genetic Algorithm, particle swarm optimization, power loss, voltage profile