Distribution network fault regionalized localization based on improved dung beetle optimization

Abstract

Aiming at the problems of accuracy degradation and slow convergence speed of traditional intelligent optimization algorithms in solving distribution network fault localization. A spiral search based multi-strategy dung beetle optimization (SMSDBO) algorithm is proposed for active distribution network fault localization. First, the hierarchical topology model of distribution network with fault tolerance is constructed, and all the segments and nodes of the distribution network are divided into different regions according to the principle of equivalence. Second, the population is initialized by logistic-Tent chaotic mapping to make the population distribution uniform, and an improved sinusoidal algorithm is added to balance the global and local search ability. Then, incorporating the spiral search strategy into the algorithm helps the algorithm to jump out of the local optimum at a later stage. Simulation experiments on distribution networks in MATLAB. Simulation results show that the combination of the SMSDBO algorithm and the hierarchical model has superior localization capabilities in single-fault, multi-fault, and information distortion fault localization. The accuracy and speed are better than the comparison algorithm and traditional model.