With the rapid development and wide application of new energy technologies, power system is facing the trend of high proportion of new energy and multi-agent interconnection. However, the dimensional explosion caused by solving the scheduling target of unit subject diversity in the whole period has become the main problem affecting the normal operation of the system. Therefore, a collaborative management method is proposed low-voltage station area based on microgrid carbon metering data. According to the dynamic cohesion of the system, the time range of optimal scheduling by the agent is divided to balance the solution dimension. At the same time, in order to solve the problems of economic operation, environmental benefit and consumption of high proportion of new energy system, a multi-objective scheduling model is constructed, which combines its dynamic cohesion and dynamic multi-objective co-evolution solution model. The results of the final example show that the model and solution method proposed can effectively reduce the computational dimension of the solution and realize the fine scheduling and management of the low-voltage station area, and improve the operation efficiency and stability of the system.