Application of multi-objective fruit fly optimisation algorithm based on population Manhattan distance in distribution network reconfiguration

In order to optimise the operation state of the distribution network in the presence of distributed generation (DG), to reduce network loss, balance load and improve power quality in the distribution system, a multi-objective fruit fly optimisation algorithm based on population Manhattan distance (pmdMOFOA) is presented. Firstly, the global and local exploration abilities of a fruit fly optimisation algorithm (FOA) are balanced by combining population Manhattan distance (PMD) and the dynamic step adjustment strategy to solve the problems of its weak local exploration ability and proneness to premature convergence. At the same time, Chebyshev chaotic mapping is introduced during position update of the fruit fly population to improve ability of fruit flies to escape the local optimum and avoid premature convergence. In addition, the external archive selection strategy is introduced to select the best individual in history to save in external archives according to the dominant relationship amongst individuals. The leader selection strategy, external archive update and maintenance strategy are proposed to generate a Pareto optimal solution set iteratively. Lastly, an optimal reconstruction scheme is determined by the fuzzy decision method. Compared with the standard FOA, the average convergence algebra of a pmdMOFOA is reduced by 44.58%. The distribution performance of non-dominated solutions of a pmdMOFOA, MOFOA, NSGA-III and MOPSO on the Pareto front is tested, and the results show that the pmdMOFOA has better diversity. Through the simulation and analysis of a typical IEEE 33-bus system with DG, load balance and voltage offset after reconfiguration are increased by 23.77% and 40.58%, respectively, and network loss is reduced by 57.22%, which verifies the effectiveness and efficiency of the proposed method.