Distribution feeder reconfiguration with refined genetic algorithm

A refined genetic algorithm for a distribution feeder reconfiguration to reduce losses is presented. The problem is optimised in a stochastic searching manner similar to that of the conventional GA. The initial population is determined by opening the switches with the lowest current in every mesh derived in the optimal power flow (OPF), with all switches closed. Solutions provided by OFF are generally the optimum or near-optimal solutions for most cases, so prematurity could occur. To avoid prematurity, the conventional crossover and mutation scheme was refined by a competition mechanism. So the dilemma of choosing a proper probability for crossover and mutation can be avoided. The two processes were also combined into one to save computation time. Tabu lists with heuristic rules were also employed in the searching process to enhance performance. The new approach provides an overall switching decision instead of a successive pattern, which tends to converge to a local optimum. Many tests were conducted and the results have shown that RGA has advantages over many other previously developed algorithms.