Considering the power consumption characteristics of multiple users helps to improve the effect of implementing demand response. Thus, this paper researches on the electricity consumption characteristics of industrial users, small and medium-sized users such as business and residents. Industrial load reduction, load transfer and demand response model for captive power plants and load aggregator models that provide both load reduction and spinning reserve are designed. These models are connected into a system model of scheduling optimization model with large-scale wind farms. With the minimum total system cost as the goal, the system is optimized by using the software Gurobi to verify different types of demand response. The results of case studies show that by stimulating different users to participate in demand response, which can effectively help system save energy and reduce emissions, improve the wind power utilization percentage, cut shaves and fill valleys.