Abstract: According to the planning problem of distributed generation (DG) in the distribution network, a multi objective optimization allocation model of DG is established by considering of schemes economic evaluated by comprehensive investment cost, the environmental protection of schemes evaluated by system loss, and the voltage stability of schemes evaluated by the voltage deviation. The distributed generation optimal allocation model is solved by using of the multi objective particle swarm algorithm, the optimized guidance strategy is adopted to improve the selection of global optimization values for the multi objective particle swarm algorithm, and the global optimization performance is effectively improved by introducing the non dominated sorting and elitist strategy into algorithm, which make algorithm quickly converge to the Pareto optimal front. Taking testing system of IEEE33 node distribution network as an example, after optimizing the distributed generation installation position and capacity, the proposed algorithm has better global convergence and searching capability compared to the results obtained with the NSGA II algorithm.