AUTHOR=Safari Ashkan , Gharehbagh Hamed Kheirandish , Nazari-Heris Morteza , Oshnoei Arman 

TITLE=DeepResTrade: a peer-to-peer LSTM-decision tree-based price prediction and blockchain-enhanced trading system for renewable energy decentralized markets

JOURNAL=Frontiers in Energy Research

VOLUME=Volume 11 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/energy-research/articles/10.3389/fenrg.2023.1275686

DOI=10.3389/fenrg.2023.1275686

ISSN=2296-598X

ABSTRACT=Intelligent predictive models play a pivotal role in P2P energy trading by accurately forecasting supply and demand variations, enabling efficient energy allocation and pricing for participants in decentralized energy markets. DeepResTrade is a research contribution that presents an advanced model for the prediction of values in renewable sustainable energy decentralized markets. This model integrates several key components, including a Peer-to-Peer (P2P) trading system, Long Short-Term Memory (LSTM) networks, Decision Tree (DT) for price prediction, and blockchain technology to enhance the trading process. With a comprehensive dataset comprising 70,084 data points encompassing market price, maximum power, minimum power, and renewable production, DeepResTrade achieves exceptional predictive performance. The model's Mean Absolute Percentage Error (MAPE) of 0.0006365308042612344% and Root Mean Square Percentage Error (RMSPE) of 0.9999988178724151% signify accuracy in capturing the complex dynamics of renewable energy markets. Moreover, DeepResTrade demonstrates its effectiveness with an RMSE of 0.01607904605926803 and MAE of 0.009125763632771632, highlighting its ability to minimize the deviation between predicted and actual values. The model further excels in explaining the variability of the actual values, as evidenced by its commendable R2 score of 0.9999988178724151. Additionally, the F1 score of [1, 1, 1], recall values of [1, 1, 1], and overall accuracy of 1 underscore DeepResTrade's precision, recall, and correctness in predicting values. The successful implementation of DeepResTrade showcases its potential to improve renewable sustainable energy markets, offering invaluable insights for real-world applications and enhancing sustainable energy trading.