Regenerative electric boilers are an important means to increase the abandonment of wind power-thermal power systems, but the current cost of wind power is still higher than the cost of thermal power. There is a contradiction between wind power consumption and the reduction of energy cost of regenerative electric boilers. Optimizing the output of wind power, thermal power units and regenerative electric boilers is of great significance for reducing energy costs and improving wind power consumption. The article establishes a mathematical model for optimal scheduling of wind power, thermal power units, and regenerative electric boilers that take into account multiple objectives. By adopting a small-scale membership function, the multi-objective function with the lowest operating cost of wind power and thermal power units and the smallest abandonment of wind power is converted into a single objective function, and the output of the wind power and thermal power units with the best membership degree is obtained under the constraints of each part. Further, a modified particle swarm optimization algorithm is proposed to solve the above model. The algorithm can jump out of the local optimal solution, and has a faster convergence speed and a higher calculation accuracy. Finally, based on the actual data of a power grid, the simulation results verify the effectiveness of the proposed method.