Multi-objective Optimisation of Distributed Generation Units in Unbalanced Distribution Systems

摘要

Due to the increased concern about the environment, distributed generation (DG) units have been widely introduced to the power system. However, DG units may have positive and negative impacts on the voltage profile and active power loss of a power system, depending on their size and location. In this paper, three algorithms namely multi-objective particle swarm optimisation (MOPSO), non-dominated sorting genetic algorithm (NSGA-II) and strength pareto evolutionary algorithm (SPEA2) were used to identify the optimum size, location and type of a DG unit in an unbalanced distribution system. The simulations were performed on the IEEE 34 Node Test Feeder System using OpenDSS and MATLAB such that the total active power loss and the voltage deviation are reduced. The effectiveness of the algorithms were evaluated based on the computation time and performance metrics such as generational distance, pure diversity and spread. It was found that all the three algorithms were suitable for the optimisation. However, NSGA-II had the lowest average computation time.