| Distribution network reconfiguration is defined as altering the topological structures of the system by operating the sectionalizing and tie switches. The distribution reconfiguration can reduce the power loss, and eliminate the overload of the lines, and improve the power quality, and restore the power supply to non-fault area in the distribution network and so on.This paper presents a power flow based algorithm for the reconfiguration of distribution networks in order to reduce the power energy losses. With all switches closed, the algorithm finds the optimal switch in each loop to be opened. Then these switches are opened to form a radial network. The algorithm finds the optimal switch to be opened when one of these switches is closed to form a one-loop network. At last, the optimal operating state of the network is obtained. The algorithm is tested by some examples, and the optimum results are quickly obtained. It demonstrates the effectiveness of the algorithm.The dissertation presents an improved evolutionary algorithm for the distribution network reconfiguration. Based on the concept of gradient and the characteristic of distribution network, the method uses acceleration operation to the excellent individual to push it to a local optimum, then, uses migration operation to keep the diversity of the population. At last, the optimum in all the local optimums is selected as the global optimum. The method overcomes the defect of producing infeasible solutions during genetic algorithm or differential evolution algorithm in the application to distribution system reconfiguration. The method is tested by two real distribution networks, and the desired result demonstrates the validity of the method. |