系统负荷作为综合能源系统的约束条件，常以单个典型日加以表征，这难以描述实际负荷波动性与随机性的特点。本文构建了计及碳交易与需求响应的多典型日优化模型。通过引入谱聚类算法获得多个典型日数据，在表征原始负荷数据波动性与随机性的同时，也基于负荷数据构成实现了数据分类，并以负荷构成为基础分别建立优化模型，引入阶梯型碳交易机制与需求响应机制，在Python语言下调用Gurobi求解器完成模型求解仿真，在此基础上对阶梯型碳交易参数的变化开展研究。结果表明，本文提出的优化模型可很好地针对全年不同的负荷特征实现灵活调度，在兼顾经济性与环保性的同时有利于清洁机组出力。

As the constraint condition of integrated energy system, system load is often characterized by a single typical day, which is difficult to describe the characteristics of actual load with fluctuation and randomness. In this paper, a multi-typical day optimization model including carbon trading and demand response is constructed. By introducing spectral clustering algorithm to obtain multiple typical daily data, While characterizing the volatility and randomness of the original load data, data classification is realized based on the load data composition, and optimization models are established based on the load composition, a stepped carbon trading mechanism and a demand response mechanism are introduced. The model is solved with Gurobi solver in Python language. The results show that the optimization strategy proposed in this paper can well realize flexible scheduling according to different load characteristics throughout the year, and is beneficial to the output of cleaning units while taking into account the economy and environmental protection.