Optimization of Multi-microgrid Operation Based on Blockchain Technology and Multi-agent Reinforcement Learning

Abstract

Blockchain technology has attracted attention in the field of electricity market due to its decentralization and fair transaction characteristics. However, the existing research lacks the quantitative analysis of the benefits brought by blockchain, and the solution efficiency is not high. In view of the above problems, the paper first proposes a local energy market (LEM) architecture based on blockchain technology, which guarantees the openness and transparency of transactions, and guarantees the privacy of participants. On this basis, a distribution market operation optimization model with multi-microgrid based on blockchain technology is proposed. Using the multi-agent reinforcement learning algorithm to solve the proposed model can effectively improve the solution efficiency. By transforming the objective function from the maximization of the overall benefit to favor of the interests of one party, it is proved that blockchain can effectively improve the overall income by guaranteeing the transaction fairness.