With the accelerated construction of the ubiquitous power Internet of things, there is an emerging understanding of the significance of aggregating the controllable load on the user side to improve the controllable capacity of the power grid and enhance the reliability of the power grid. This paper explores the type of user-side controllable load and focuses on demand-side response strategy. Firstly, from the perspective of system operators, a mathematical model is constructed to balance the real-time demand response of large industrial users, load concentrators, end-users participating in demand response and power generation operators'' up-regulated output, also, the objective function of minimum adjustment cost of system operators is established. After that put forward the demand response Stackelberg''s mathematical model, system operators as a leader, demand responser to the market as followers, system operators to provide incentives the followers, the market participant to make decisions based on incentive compatibility principle, at the same time, the system adjustment cost minimum. Finally, based on a real example, the decentralized decision-making algorithm of particle swarm optimization was applied to consider the power boundary constraint of demand response, set load priority, and solve the demand response strategy. The simulation results of the example verified the feasibility and effectiveness of the established model and method.