Abstract: As an important interactive resource between generation side and demand side, demand response can effectively regulate the distribution of load demand to achieve energy-saving and emission-reduction and improve the system wind power capacity. Based on this, this paper considers the demand response in the environmental economic and proposes a day-ahead two-stage dispatching model under smart grid. The first stage is day-ahead user interaction stage, in which the next day load distribution is adjusted by time-of-use price leverage guiding the user to take rational power consumption and the optimal load curve and time-of-use price is determined by considering the load level and user satisfaction index. The second stage is day-ahead dispatching stage, in which the economic emission dispatch model is established based on chance-constrained programming for wind power randomness and this model is transformed into a deterministic model by using wind power distribution function. We propose an improved multi-objective particle swarm optimization algorithm by introducing the diversity index, random black hole theory and the multi-targeted search mechanism, and adopt technique for order preference by similarity to ideal solution（TOPSIS） method to sort the Pareto frontier individual to help the dispatcher to make scientific decision. The simulation results of the improved 10 machine system verify the validity and rationality of the model and method.

Key words: wind power, demand response, time-of-use price, customer's satisfaction, economic emission dispatch, technique for order preference by similarity to ideal solution（TOPSIS）