Distribution Network Reconfiguration Based on Simulated Annealing Immune Algorithm

Abstract

A distribution network reconfiguration method based on simulated annealing immune (SAI) algorithm is proposed in this paper. According to the characteristics of reconfiguration, loop encoding is used and the public switches between loops are disposed in advance, so the individuals in evolution turn to be feasible. Besides, the historic and present optimal individuals are combined to construct vaccine dynamically, and not only the speed of evolution searching but also the individual performance can be improved by inoculation. On the other hand, the selection operator based on Boltzmann annealing method, high frequency mutation and immune supplement are used to keep the population diversity. Through the improvements, the early maturity is avoided effectively. Reconfiguration results show the prompt convergence and effectiveness of proposed method.