Design of Peer-to-Peer Transactions for Multi-microgrids Based on Non-cooperative Game Theory and Distributed Machine Learning

• Published in: 2022 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia) pp. 984 - 990
• Main Authors: Li, Jifeng, Pan, Feng, Wang, Yanbo, He, Xingtang, Liang, Xianming, Wu, Jun
• Format: Conference Proceeding
• Language: English
• Published: IEEE 08.07.2022
• Subjects: distributed machine learning method; multi-microgrids; non-cooperative game; peer-to-peer trading
• DOI: 10.1109/ICPSAsia55496.2022.9949835

In response to the lack of further game relationship analyses and gaps in the application of artificial intelligence methods represented by machine learning in peer-to-peer distributed trading, this paper proposed a multi-microgrids peer-to-peer trading method based on non-cooperative game theory and a distributed machine learning algorithm. First, the autonomous scheduling and peer-to-peer game trading models of the peer-to-peer market trading subject, i.e., the microgrid, were constructed. Second, the principle of the elastic average stochastic gradient descent algorithm and the distributed machine learning framework based on the algorithm were introduced. Finally, the effectiveness and applicability of the theories and methods proposed in this paper were verified through practical arithmetic simulations. This paper explored the applicability of a distributed machine learning algorithm in peer-to-peer market trading.