In order to improve the consumption level of intermittent new energy in active distribution network （ADN）， this paper presents an economic optimization model for ADN energy dispatching based on multi-agent dynamic topology reconfiguration. The model takes the minimum reconstruction power loss， the maximum utilization efficiency of distribution new energy sources and the maximum profit of distribution companies as the decision-making conditions， and the economic maximization is achieved through the energy scheduling of each link in ADN. ADN energy dispatching is implemented by three-layer time-division and space-distribution control strategies using multi-agent system （MAS）. The chaotic quantum particle swarm optimization algorithm is used to solve the ADN energy optimal dispatching equations to realize the benefit optimization from the individual to the whole. Finally， taking four typical scenes of IEEE 69 node distribution network as examples， ADN energy dispatching is simulated. The effectiveness of the model is verified while the economic optimal solution is obtained.