随着电动汽车、储能设备、柔性负荷等可控负荷参与系统调度运行的技术条件日益成熟，利用可控负荷跟踪不可控电源出力的负荷调度成为提高含间歇性新能源电力系统运行效益和消纳能力的有效方法。本课题拟从一个双互动协调角度，即同时考虑源/荷的独立互动以及直接/间接负荷控制模式的协调融合，利用不确定性分析和多模态学习进行负荷调度方法研究。①分析可控负荷的控制特性及多时间尺度响应特性，提出单一控制模式下考虑间歇性能源随机性的负荷调度模型，揭示调度运行中源/荷间的互动关系；②分析多个负荷控制模式之间的交互影响，设计基于多模态学习的负荷控制模式关联分析方法，并构建逼近运行实际的负荷调度模型；③提出针对单一控制模式下负荷调度模型的多重调度轨迹解推求方法和多控制模式并存下负荷调度模型的快速求解方法。相关研究成果可以提高源/荷两侧调峰资源的精细化协调控制，对提升我国新能源接入能力和系统运行精细化水平提供理论和技术支持。

With the growing maturity of techonology conditions for controllable load（electric cars, energy storage and flexible load）participating in dispatching operation in power system, using controllable load to track uncontrollable power output, the load dispatch method is a good way to improve the efficiency and ability of accommodation of power system with intermittent new energy. This issue intends to coordinate the interaction of source and load as well as the direct and indirect load control mode. The research is based on Equilibrium Game theory and multimodal learning. Research topics include: ① constructing cost function, load response model and shifting peak predictive model based on the ability to control the characteristics of different types of controllable load, proposing load dispatch model considering intermittent new energy under single control mode and the interaction relationship between source and load. ② analyzing the interaction of multiple load control modes, designing interactive load dispatch method considering multi-control mode and constructing load dispatch model that is approximate to the real operation. ③ proposing a method to solve the multiple scheduling trajectory in load dispatch model under single control mode and a quick solution of load dispatch model under multiple control mode. Related research can improve the coordination of source and load interaction and offer theoretical and technical support in the area of new energy.