In a microgrid system

the distributed generator has the characteristics of being intermittent and fluctuating

and the load is variable. The uncertainty of the distributed generator and the load demand will lead to the uncertainty of the microgrid energy optimization results. In order to solve the above problems

an improved model predictive control(MPC)strategy for two-layer microgrid energy optimization with multi-time scale is proposed. As some model parameters are fixed in the traditional MPC optimization

it is hard to timely deal with the emergency disturbance in the system; and the single optimization cycle time of traditional MPC is limited

thus it is hard to deal with some complex constraints which are related to time or affect the overall optimization results. An adaptive improved MPC strategy is proposed according to the uncertainty of microgrid and complicated constraints to adapt to the microgrid characteristics of equipment switching flexibility and power generation affected by the outside world. It is shown that the robustness of the system and the accuracy of optimization are guaranteed. The allocation of future energy and load scheduling is optimized according to the day-ahead plan. By taking the results of the day-ahead optimization as reference

the real-time energy optimization is undertaken based on MPC

so as to make the optimization goal better

and improve the accuracy of the results. Finally

the simulation results of MATLAB demonstrate the applicability and accuracy of the proposed method.