In order to ensure stable operation of each component in the microgrid and improve its economic efficiency, firstly, the multi-agent technology is used to coordinate the demand side load, the battery and the generator side micro source, so that the information between agents in the microgrid can be timely output to reduce the system failure. Secondly, a battery charging and discharging control strategy is designed based on the real-time electricity price microgrid mechanism connected to the grid. The output power characteristics of the battery are analyzed to coordinate the output of each micro source at different times of the day. Meanwhile, the Multi-Agent Chaotic Particle Swarm Optimization (MACPSO) algorithm is adopted to solve this proposed model. Through the simulation contrastive analysis, it is verified that the microgrid based on multi-agent coordinated control can effectively improve the economic operation efficiency. This work is supported by National Natural Science Foundation of China (No. 51407104 and No. 61603212).