Probabilistic distribution network expansion planning with multi-objective Memetic Algorithm

This paper presents an efficient method for probabilistic distribution network expansion planning with Multi-objective Memetic Algorithm (MOMA) that combines Multi-objective GA with LS (local Search). Recently, the emergence of the deregulated and competitive power markets makes distribution networks more complicated. In this paper, a Monte-Carlo-Simulation (MCS) based method is proposed for handling the uncertainty of distribution network expansion planning. To improve the performance of MCS, the proposed method considers the correlation between the nodal specified values to evaluate more realistic results. Also, this paper focuses on the multi-objective meta-heuristic technique that evaluates a set of the Pareto solutions systematically. Distribution network expansion planning has a set of objective functions such as the installation cost of new feeders, new distribution substations and distribution generators, the network loss, the reliability index, etc. As a multi-objective optimization technique, this paper makes use of Controlled-NSGA2 of multi-objective meta-heuristics (MOMH) that modifies NSGA2 to enhance the diversity of solution sets. Furthermore, this paper introduces the idea of Memetic algorithm (MA) into controlled-NSGA2 to improve the solution quality. LS in MA plays a key role to improve the solution evaluated by GA. The proposed method is successfully applied to a sample system.