In order to improve the performance of intelligent algorithms and the optimization efficiency;the dynamicadaptive particle swarm optimization（DAPSO）is presented in this paper to search the best scheme of network recon-figuration with distributed generation thus to meet the network radiation and connectivity constraints;and the recon-figuration of distribution is a discrete variable optimization problem. The DAPSO can adaptive change inertia weightand the speed of the particles. It maintains the global development and local exploration of dynamic balance thus canenhance the optimization performance of the algorithm. The"loop disconnect"method is used to solve the problem ofreconstruction. The network is kept in radiated and it can ensure network connectivity at the same time after configu-ration. The simulation results based on IEEE33 nodes system and PG & E69 nodes system show that DAPSO algo-rithm has the advantages of fast convergence speed;strong global optimization ability and great stability. Its optimalreconstruction scheme can effectively reduce the network loss and improve the voltage level. It’s better than othermethods and has great practical value.