Abstract: During the process of distribution network reconfiguration by intelligent optimization algorithm using basic ring matrix coding the disordered solution space is usually utilized, however there are many local peaks in the solution space so that it makes the intelligent optimization algorithm hard to full play its own advantages, besides it consumes too much time and it is difficult to search out the optimal solution. In allusion to above-mentioned defects, an ordered ring network coding scheme was proposed and the improved gray wolf algorithm was used to search out the method to reconfigure distribution network containing distributed generation (DG). Firstly, the elements of the basic ring matrix wee arranged according to the order of branches, and then the heuristic rules were used to preliminary find the better solution and this better solution was compared with the alpha wolf in the initial wolf pack, and the better solution was taken as the new alpha wolf. Secondly, the gamma wolf was led in to surround the alpha wolf for optimization to ensure the population diversity of the wolf pack meanwhile to improve its local searching ability. Finally, the improved grey wolf algorithm was used to conduct the simulation in the modified IEEE 33-bus distribution network and the Taipower84 distribution system. Simulation results show that using the proposed reconfiguration method the network loss is effectively reduced and the lowest voltage in the system is improved, thus, the proposed reconfiguration method is effective and feasible, the algorithm is simple and fast, and the obtained results are better.