Based on the user′s response mechanism and price elasticity theory in the peak-to-valley time-of-use price, the user will adjust the power consumption with the change of the electricity price. Therefore, the user will voluntarily adjust the power consumption plan through the electricity price change, thereby reducing the load curve. The peak-valleydifference reduces the cost of each side of "source", "grid"and "load" and increases the income. In this paper, the fuzzy clustering algorithm is used to classify the daily load, and the real-time electricity price model is established according to the peak-middle-valley membership degree of each hour. Based on the relevant policies after the new electricchange, the mathematical model of the power supply side and the grid side is established. The "source", "grid", and "load" sides of the revenue are the optimization model with the largest gain on the grid side and the minimum peak-valley difference as the objective functions, and the load side user electricity cost does not rise as the constraint. The particle swarm optimization algorithm is used. Solve and verify the model′s narrow peak-valley difference and increase the effectiveness of the gain on the Matlab simulation platform.