Aiming at the shortcomings that the traditional standard particle swarm optimization (PSO) tends to fall into the local optimum and the differential evolution algorithm (DE) has a slow convergence rate in the later stage,a differential particle swarm optimization algorithm (DEPSO) is proposed to optimize both the convergence speed and the global Optimum ability,and applied to distribution network substation planning.Through the combination of Voronoi diagram in the mathematical model of substation site selection to determine the substation power supply range and planning capacity,and then verify the substation actual load rate,simplify calculations and improve search efficiency.The long-term planning example of a city city verified that the algorithm is correct and effective,which can meet the planning requirements of urban distribution network.