Multi-entity Behaviors in Electricity-Carbon-Green Certificate Coupled Markets Based on Multi-agent Reinforcement Learning

Establishing a national carbon emission trading market and green certificate market is one of the key strategies for China to achieve its "dual carbon" goals. However, existing research predominantly analyzes the market coupling relationships from an economic perspective, overlooking the impact of power network physical constraints and the uncertainty of renewable energy output on the coordinated optimization of markets. Additionally, the scenario of power consumers participating in the carbon market has not been considered. To address this limitation, a bi-level optimization model for the coupled electricity-carbon-green certificate market based on physical network nodes is proposed to analyzes the behaviors of market entities and the changes in coupling mechanisms under the context of carbon market expansion. In the model, a decision-making mechanism is introduced for power consumers participating in the carbon market based on the physical topology of the power grid, and by incorporating the offset rules between green certificates and carbon allowances, the impact of transmission line congestion on the decision-making of market entities is explored. The actual output data of new energy units in the Mongolia region is used to verify the rationality and effectiveness of the proposed model. The results show that the participation of electricity users in the carbon market can significantly increase the overall returns of the coupled market, and line congestion has a significant impact on the behaviors of market entities and market revenues; in the context of abundant carbon quotas, introducing the carbon credit offset mechanism can further optimize the coupled market efficiency.