Abstract:Through the integrated management of distributed power sources, the combined cooling, heating and power（CCHP）-based microgrid not only improves the efficiency of renewable energy utilization and reduces the environmental pollution, but also ensures the stability of the system. Based on the uncertainty of each micro-source and the complexity of microgrid structure, a comprehensive real-time power demand optimization model based on the pre-day initial load demand and combined with electricity price incentive and user-side electricity comfort is proposed, in which the shift of load from peak to non-peak is realized. And the multi-objective optimization algorithm based on genetic algorithm is adopted to prove that the proposed model can make the load demand curve obviously smooth and the output of each micro-source is relatively balanced, which proves the proposed model is scientific and effective.