An efficient multi-objective memetic algorithm for uncertainties in distribution network expansion planning

This paper proposes an efficient multi-objective memetic algorithm for distribution network expansion planning (DNEP). It may be expressed as a complex multi-objective optimization problem as well as combinatorial optimization one. Recently, deregulated and competitive power markets brought about the uncertainty of distribution systems. There are correlations between the nodal specific values. A Monte-Carlo-simulation-based method is proposed for handling the uncertainty and correlation. Also, this paper focuses of multi-objective meta-heuristics (MOMH). It obtains many final solution candidates at the same time. Also, MOMH has specific strategies in terms of high accuracy and diversity of the set. This paper makes use of the improved strength Pareto evolutionary algorithm (SPEA2) and the controlled fast elitist non-dominated sorting genetic algorithm (CNSGA2) that are well-known for more effective methods in MOMH. As memetic algorithm (MA) that integrates meta-heuristic with local search (LS). This paper combines SPEA2 and CNSGA2 with random-multistart local search (RMSLS) for improving the solution set. This is called memetic algorithm (MA). RMSLS plays a key role to provide many solution candidates. The proposed method has advantage to keep high accuracy and diversity of solution sets in MOMH. The proposed method is successfully applied to a sample system.