Data-driven Pricing Strategy of New Energy Stations in The Spot Market Environment

With the continuous expansion of the proportion and scale of renewable energy access, the participation of renewable energy in electricity market trading and multi-dimensional collaborative consumption are becoming the trend of renewable energy grid connection. The impact of spot clearing prices on the trading revenue of new energy stations is particularly important. To help new energy stations better participate in the electricity spot market, a pricing strategy formulation method based on Deep Reinforcement Learning (DRL) considering daily rolling trading and day-ahead spot trading of new energy stations is proposed. Firstly, historical data on the power generation of new energy stations, daily clearing electricity prices, upper and lower limits of quotations, and plant power consumption rates are collected. Secondly, the state space, action space, and reward function of the deep reinforcement learning agent is set, and the agent selects a pricing strategy in the action space and evaluates the $\mathbf{Q}$-value. Finally, the $\mathbf{Q}$-value and neural network parameters are updated, and the pricing strategy are quickly obtained for new energy stations based on the operating environment data of the electricity market during the online phase. Through comprehensive case studies, it is shown that the proposed method for formulating trading pricing strategies for new energy stations can effectively help them obtain higher returns in complex market environments.