Design of a multiagent simulation and game theory framework for intelligent decision-making in electricity trading

In the context of the dual carbon target, the electricity market presents characteristics such as high penetration rate of renewable energy, diversified trading entities, and dynamic trading rules. Traditional decision-making models rely on human experience and are no longer suitable for complex market environments; Traditional power trading systems generally suffer from weak multi-party interaction and cooperation capabilities, disconnection between game mechanisms and actual trading rules, and limited decision-making adaptability in dynamic environments. The research content of this article is as follows: Firstly, a multiagent system with a four layer architecture of "subject layer game layer decision layer simulation layer" is constructed, clarifying the responsibilities and interaction interfaces of participants such as power generation enterprises, power sales companies, and power users; Secondly, by integrating non cooperative game theory and cooperative game theory, a mixed game mechanism is designed by embedding clearing rules and auxiliary service trading logic in the electricity spot market; Finally, a reinforcement learning and fuzzy decision-making module is proposed to achieve intelligent decision-making optimization of the project. The experimental results show that the delay of Frame-A gradually increased from 0.38s at 0% fluctuation to 0.72s at 40% fluctuation (a gentle rise). The delay of Frame-B has reached 1.15s (over 1 second) when fluctuating at 0%, increased to 1.40s when fluctuating at 15%, and further increased to 1.85s (sharply rising) when fluctuating at 40%, verifying its efficiency and practicality.