Abstract: Combined heat and power (CHP) microgrid uses electricity and natural gas as sources of heat energy， which has the characteristics of economy and environmental protection. It can effectively solve the problem of comprehensive power and heat supply， and has broad application prospects in industrial parks. Considering the randomness of renewable energy in CHP microgrid and the error of load forecasting， the accuracy of energy optimization will be reduced. And there are differences between electric and thermal loads， and the economy of CHP microgrid with high thermoelectric coupling is poor. This paper proposes a multi-time-scale energy optimization method based on model predictive control (MPC). The allocation energy and storage scheduling is optimized according to the day-ahead plan. Applying the MPC algorithm， the real-time energy optimization is planned in the day， and the optimization results are improved due to the feedback design of heat loss in the system. Finally， in the case simulation in MATLAB， the applicability and the accuracy of the proposed method is verified， and the operation cost of the system is reduced. The simulation results show that the method achieves the effect of thermoelectric decoupling and peak shaving and valley filling by energy storage， meanwhile improving the utilization of renewable energy.

Key words:  industrial park, microgrid, combined heat and power (CHP), multi-time scale, energy optimization, model predictive control (MPC)