Optimasi Penempatan dan Kapasitas Distributed Generation Menggunakan Metode Hybrid GA-PSO untuk Mengurangi Rugi-Rugi Daya dan Meningkatkan Profil Tegangan

Abstract

Radial distribution systems often suffer from high power losses and voltage deviations, motivating careful siting and sizing of distributed generation (DG). This study analyzes loss and voltage profiles on the IEEE 33-bus test system using the Backward–Forward Sweep power-flow method, and then optimizes DG locations and capacities via theoretical placement as well as metaheuristics: Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and a hybrid GA–PSO. Without DG, active and reactive power losses are 202.682 kW and 135.2372 kVAr, respectively, with several buses below 0.95 p.u. For a single-DG optimization, GA and hybrid GA–PSO yield identical results of 109.65 kW and 76.408 kVAr, while PSO gives 109.682 kW and 76.507 kVAr, all improving the voltage profile. For the five-DG configuration, GA achieves 74.724 kW and 51.229 kVAr, PSO 79.424 kW and 54.322 kVAr, and the hybrid GA–PSO performs best with 70.978 kW and 47.898 kVAr, producing the flattest profile near 1.0 p.u. These results confirm that all metaheuristic methods outperform the base case, with the hybrid GA–PSO proving most effective for multi-DG placement in simultaneously reducing system losses and enhancing voltage stability.