Peer to Peer Energy Transaction Market Prediction in Smart Grids using Blockchain and LSTM

Abstract

Predicting future demand for distributed energy sources is difficult because of the energy generations is unpredictability. We proposed an auto executable blockchain-based peer-to-peer energy transaction market with Long ShortTerm Memory (LSTM) neural network for energy trading and predictions. Our proposed energy transaction market and trading strategy are built via smart contracts inside the blockchain. The energy trading process in the market is auto executable and does not need a long waiting time as in the conventional bidding strategy. LSTM is integrated into the market for predicting future energy usage. Experimental results show that the proposed blockchain-based peer-to-peer energy transaction market can earn more profit than the traditional method.