Modern intelligence optimization methods for distribution network reconfiguration have been introduced in detail.Their strong and weak points have been discussed.Simulated annealing algorithm can work out optimal or near optimal solution independent of initial solution

however

they are time-consuming and sensitive to parameters of algorithms.Genetic algorithm cannot be constrained by the features of the object function and the performances of search are efficient owing to parallel operation

but they sometimes converge to local optimal solution.Tabu search algorithm can avoid local optimal solution for their flexible memory function

however

they are sensitive to initial solution and the search operations are point-to-point.The algorithm based on artificial neural network can find out optimal solution quickly owing to no calculation of flow

but the results depend on the training sets.