Power generation side member dynamic quotation decision-making method, device and equipment

The invention discloses a power generation side member dynamic quotation decision-making method, device and equipment, and the method comprises the steps: S1, generating a reinforcement learning initial strategy space of each power generation member, and carrying out the initialization of the selection probability of each quotation strategy; S2, selecting a quotation strategy with the maximum selection probability; S3, utilizing the clearing model to obtain a market clearing result; S4, generating a local optimal strategy space based on a diagonalization optimization mechanism; S5, calculatingan evaluation function value of each quotation strategy through a reinforcement learning method, and correcting a selection probability of the quotation strategy; S6, when the reinforcement learningtermination condition is not met, executing the step S2 again; and S7, when it is judged that the evaluation function value meets the reinforcement learning termination condition, outputting an optimal quotation strategy. Acco