Blockchain and Federated Learning Based Bidding Applications in Power Markets.

Currently, power generators are facing an imperfectly competitive power market with relatively limited information, and they can increase their own economic returns by raising their quotations through game theory, which leads to an inflated price in the wholesale power market. In this paper, we propose a bidding application for the wholesale power market called Federated Learning Power Quotation System (FedLPQS), which combines federated learning and blockchain to consider the characteristics of the future power trading market of "Liberalizing both ends". FedLPQS using federated learning aggregates cost data of each power plant to achieve a certain degree of transparency and records the data in a blockchain platform based on the consortium chain to reduce the price of power traded in the wholesale power market. [ABSTRACT FROM AUTHOR]