Credit-based energy trading system using blockchain and machine learning.

In peer-to-peer (P2P) energy trading, members locally trade energy. Blockchain-based systems are employed for the above trading. These systems are limited in speed because of the time required in the consensus process to audit and verify transactions. Further energy is traded using auction, which also consumes time. In this paper, we propose a blockchain and machine learning-based system to speed up trading. This system uses Byzantine Fault Tolerance method to reduce consensus time and machine learning to reduce auction time. In addition, the proposed system provides loans to cash-deficient buyers thereby enabling them to purchase energy. Machine learning is used to arrive at optimal loan rates for buyers. All these claims have been verified by the simulation results. Simulation results show that the new mechanism finds the optimal value of the trader's quotes prices quickly by altering the value of the profit margin, and the new loan system increases the number of transactions in this blockchain-based system. Hence, the proposed system will promote P2P trading by making it faster and enabling greater participation of cash-deficient buyers in the trading. [ABSTRACT FROM AUTHOR]