Considering the effect of distributed power on grid structure of the distribution network and without decided capacity and location of the distributed power, the approach of inner and outer layers bi-level programming is adopted for the integrated optimization of the distribution network circuit structure, and the location and capacity of the distributed power. The government subsidy policy is introduced into the objective function, which encourages the use of clean energy, so as to build an objective function module which saves the most economic cost. For the inner layer, with the introduction of the distributed power position and capacity, the tissue membrane computing communication rule is introduced to the genetic algorithm, to enhance the ability of global optimization and computational capabilities, so as to achieve the expansion planning of the network structure; for the outer layer, the new hybrid particle swarm optimization algorithm is adapted to determine the location and capacity of the distributed power. Examples show that the algorithm applied to distribution network expansion planning including distributed power saves the cost, improves the reliability of electricity, and demonstrates the effectiveness of the algorithm.