Combined cooling, heating and power (CCHP) system integrated with multiple renewable energies can improve the environmental performance while increasing the dependency on the national grid due to the multiple uncertainties. This paper proposes a multi-objective stochastic multi-scenario optimization method for the optimal capacity of a CCHP system integrated wind, solar and geothermal energy considering its energy supply independence, environmental impact, economic performance and energy efficiency. The spatiotemporal multiple uncertainties in wind velocity, solar irradiation and multiple loads are characterized by the stochastic multi-scenarios generated by the stochastic hierarchy scenario-generation method. The hybrid system's flexibility is evaluated by both the grid integration level and net interaction level. The environmental performance is assessed by both the carbon emission reduction rate and renewable energy penetration. The economic and energy performances are characterized by the annual cost-saving rate and primary energy-saving rate, respectively. The multi-objective stochastic optimal design method is formulated and solved using non-dominated sorting genetic algorithm II. The case study results show that there are slight deviations of objectives between the stochastic multi-scenario and traditional optimizations, while the former can save 58.69% computation time. Moreover, the installed capacity increase of the electrical energy storage to 870kWh can reduce the system's dependency on the national grid and the net interaction level to 3.74% while deteriorating the economic performance and dropping the annual cost-saving rate to −217.37%. Inversely, the installed capacity increase of renewable energy generators can enhance economic and environmental performances while worsening the net interaction level.

结合多种可再生能源的冷，热和电（CCHP）组合系统可以改善环境绩效，同时由于存在多种不确定性，因此增加了对国家电网的依赖。针对风能，太阳能和地热能的CCHP系统的能量供应独立性，环境影响，经济效益和能源效率，提出了一种多目标随​​机多场景优化方法。风速，太阳辐射和多重负荷的时空多重不确定性的特征在于随机层次情景生成方法生成的随机多重情景。混合系统的灵活性通过网格集成级别和网络交互级别进行评估。环境绩效通过碳减排率和可再生能源渗透率进行评估。经济和能源绩效分别以年成本节约率和一次能源节约率为特征。采用非支配排序遗传算法II，提出并求解了多目标随机最优设计方法。案例研究结果表明，随机多方案与传统优化方案之间的目标偏差很小，而前者可节省58.69％的计算时间。此外，将蓄能器的装机容量增加到870kWh，可以将系统对国家电网的依赖性和净互动水平降低到3.74％，同时使经济表现恶化，并将年度成本节省率降至-217.37％。