基于强化学习模型的需求侧用户智能报价策略研究

Research on Demand Side Loads Intelligent Bidding Strategy Based on Reinforcement Learning Model

随着中国电力市场建设的提速进行,需求侧用户将成为市场交易主体的重要组成部分,亟需实用且智能的报价策略。结合不同需求侧用户的调节特性,针对连续调节型和离散调节型负荷建立了详细的分段调节成本模型,并形成一系列不同利润率报价方案。在此基础上,考虑市场需求及历史交易情况等因素,提出了一种基于自适应强化学习模型的智能报价策略,并阐述了报价策略的确定流程。最后,通过算例仿真验证了该报价策略简单、易操作,可进一步提升需求侧用户的报价决策能力。

With the rapid construction of power market in China, a more practical and intelligent bidding strategy is needed for demand-side loads when they become the market participants. Based on the different regulation characteristics, a piecewise regulation cost model is established for continuous or discrete regulation demand loads. And a series of different profit rate pricing schemes are formed. Meanwhile, considering the factors besides market demand and historical data, the paper proposes the intelligent bidding strategy based on an adaptive reinforcement learning model, and gives the flow for the process of the bidding strategy. Finally, the simulation results show that the bidding strategy is simple and easy to operate, which will further improve the pricing decision-making ability of the demand side loads.