Orderly Charging Price Mechanism for Vehicle Network Interaction Using Transformer-CNN

China’s electric vehicle ownership is rising year by year, becoming a potential resource that the power grid can dispatch. To better utilize V2G technology to serve the grid, it is necessary to establish a relevant market, which cannot be separated from government support and subsidies. Previous V2G research focused on multi-scenario applications, based on EV simulation, with more assumptions and simplifications on EV user behavior, which is difficult to reflect real user habits. Predicting EV charging order distribution based on charging pile historical data would be more accurate and credible. In this paper, firstly, based on the EV vehicle-network interaction mechanism, we propose a model to formulate the upper and lower limits of subsidized price considering the cost reduction price and narrowing the peak-valley difference of the power grid. Then it establishes Transformer and CNN based, predicting order distribution for charging and discharging model solving. Finally, two different operating habits of private cars and buses are brought into the vehicle data to verify the effectiveness of the price mechanism proposed in this paper. The strategy provides a reference for real-time incentive adjustment to orderly charging for vehicle-network interaction, which contributes to the long-term construction and operation of the V2G market.