Abstract: The optimal operation of the microgrid system can effectively promote the implementation of the dual “carbon strategy”. However, the dynamic changes of the load and the multi-objective characteristics of the operation are not considered during the operation of the microgrid, which limits the optimization of the system operation and the consumption capacity of renewable energy. Aiming at this problem, a multi-objective optimal operation model is established to minimize the total operating cost of microgrid and maximize the renewable energy consumption, which is composed of operating cost and cost of environmental treatment. Aiming at the dynamic change of power grid load, Holt-winters model is used to predict the load change of microgrid system. Finally, the improved multi-objective particle swarm optimization algorithm is used to solve the microgrid model. In order to verify the feasibility of the algorithm, the high proportion of new energy HRP-38 standard test system and the microgrid data in East China are studied. The example shows that the wind power utilization rate of the two types of microgrid systems is increased by 8.13% and 5.23%, respectively, the photovoltaic utilization rate is increased by 8.59% and 6.49%, respectively, and the total operating cost of the system is reduced by 8.29% and 8.89%, respectively. Therefore, this proposed method can not only effectively improve the system’s ability to absorb renewable energy, but also reduce the total operating cost of the system, while maximizing environmental benefits, which has guiding significance for microgrid operation.