Power loss reduction through distribution network reconfiguration using feasibility-preserving simulated annealing

Abstract

The paper presents a novel algorithm for solving distribution network reconfiguration problem in the context of power loss reduction and voltage profile improvement. The proposed methodology exploits a sequential approach, which is an enhanced simulated annealing routine including both local (deterministic) and random modifications of the network. Local operations search for the best network alterations of individual connections, oriented towards objective function improvement. Occasional random changes are accepted depending on the corresponding objective function value and with probability decreasing towards the end of the optimization run. All algorithm operations are designed to preserved radial configuration of the network. Extensive numerical experiments performed for the standard 33- and 69-bus systems demonstrate that our algorithm outperforms recently published state-of-the-art metaheuristic algorithms with respect to solution repeatability and the computational cost.