电力市场、碳排放权交易市场以及核证自愿减排市场耦合下发电商竞价策略

Bidding Strategy of Power Generators Under the Linkage of Electricity Market,Carbon Emission Trading Market,and Certified Emission Reduction Market

碳-电耦合是推动我国能源电力行业绿色低碳转型、助力实现“双碳”目标的重要保障。设计了电力市场、碳排放权交易市场(carbon emission trading market,CET)以及核证自愿减排市场(certified emission reduction market,CER)的耦合框架,构建了发电商协同参与电力市场、CET市场以及CER市场的竞价策略优化模型。针对传统博弈论方法对不完全信息以及复杂多元主体交互性解释不足等问题,采用深度确定性策略梯度算法进行求解,同时克服了传统多智能体强化学习算法离散低维状态空间、动作空间以及收敛不稳定等问题。以4个典型发电商进行算例分析,研究了CET市场和CER市场引入后对发电商竞价策略的影响。围绕碳市场价格、碳排放基准值以及核证自愿减排抵消比例开展敏感性分析并开展算法收敛性验证,体现了所设计的碳-电市场耦合下发电商竞价策略优化模型的有效性和优越性。

Carbon-electricity coupling is an important guarantee for promoting the green and low-carbon transformation of Chinas energy and power industry and helping to achieve the"dual carbon"goals.A coupling framework for the electricity market,carbon emission trading market(CET),and certified emission reduction market(CER)is designed and the optimization model for bidding strategies for power generators is constructed to collaborate in the electricity market,CET market,and CER market.In response to the problems of incomplete information and insufficient explanation of complex multi-agent interactions in traditional game theory methods,a deep deterministic policy gradient algorithm is adopted to solve the aforementioned problem.In addition,it can also overcome the problems of discrete low dimensional state space,action space and convergence instability in traditional multi-agent reinforcement learning algorithms.Taking four typical power generators as examples,this study investigates the impact of the introduction of CET and CER markets on the bidding strategies of power generators.Sensitivity analysis and algorithm convergence verification are then conducted around carbon market prices,carbon emission benchmark values,and certified emission reduction offset ratios,demonstrating the effectiveness and superiority of the designed bidding strategy model for power generators under the coupling of carbonelectricity market.