Aiming at the problem of multiagent interest game planning in the open environment of electricity market, a hierarchical planning model is proposed for Active Distribution Network with the coordination and interaction of "sourcenetworkload" multi agents based on the game theory, and the iterative search method and improved particle swarm optimization algorithm is applied to get the solution. The upper layer of the model considers the respective decisions of the source, network and load agents, and aims at achieving the equilibrium in the game mode for each interest body. The lower layer considers the DG(Distributed generation) reduction and OLTC(Onload Tap Changer)adjustment strategies from the active management level. The target of the lower layer is minimizing the expected value of annual load cutting of DG. In the upper layer, the iterative search algorithm is utilized to search the equilibrium solution of multiagents, and the improved particle swarm optimization algorithm is employed in the optimization process of each agent in the upper layer and the optimization process in the lower layer.Finally, an improved IEEE 33node example is simulated to verify the feasibility of the model. It also gives ideas and suggestions for the ADN planning game problem with multidecisions in the open environment of electricity market.