Power distribution network reconfiguration: hybrid PSO-Tabu Search approach for computational effort reduction

Abstract: A heuristic algorithm inspired by a descendant generating tree is proposed to delimit the search space of the distribution network reconfiguration problem. The sections with the greatest losses from the base configuration are given priority for reconfiguration, and the heuristic is combined with the PSO (Particle Swarm Optimization) and Tabu Search (BT) metaheuristics to explore and refine the solution space. The results show a reduction of up to 90% in computational effort compared to traditional approaches via a single metaheuristic to reach the optimal solution, making reconfiguration feasible in large networks. The methodology is validated for networks of 14, 33 and 84 bars.