RECreate: A Blockchain-Based Marketplace for Renewable Energy Credits with Intelligent Peer-to-Peer Trading Using Model-Based Multi-Agent Reinforcement Learning

Authors

  • Saachi Peswani Department of Computer Engineering, Sardar Patel Institute of Technology, Mumbai, India
  • Khushi Parekh Department of Computer Engineering, Sardar Patel Institute of Technology, Mumbai, India
  • Abhishek Sharma Department of Computer Engineering, Sardar Patel Institute of Technology, Mumbai, India
  • Param Gogia Department of Computer Engineering, Sardar Patel Institute of Technology, Mumbai, India
  • Sunil Ghane Department of Computer Engineering, Sardar Patel Institute of Technology, Mumbai, India

Abstract

Renewable Energy Credit (REC) markets suffer from high transaction costs (3–5%), slow settlement (2–3 days), and exclusion of small-scale producers. This paper presents RECreate, a blockchain-based marketplace integrating business-to-business (B2B) REC trading with intelligent peer-to-peer (P2P) energy exchange. The B2B component lever- ages Polygon blockchain smart contracts with zero-knowledge proofs for privacy-preserving verification. The P2P component introduces MB-MASAC (Model-Based Multi-Agent Soft Actor-Critic), combining Temporal Fusion Transformer (TFT) forecasting with differential attention for proactive multi-agent coordination. Over 7,200 simulated days, RECreate achieves 99% reduction in settlement time, 93% decrease in verification costs, and 80% reduction in transaction fees. For P2P trading, MB-MASAC achieves 42.7% cost reduction ($755 annual savings per household), 15.1% lower battery volatility extending lifespan by 20–25%, and 8.4% MAPE for 24-hour forecasting.

Downloads

Download data is not yet available.

Downloads

Published

07-05-2026

Issue

Section

Articles

How to Cite

[1]
S. Peswani, K. Parekh, A. Sharma, P. Gogia, and S. Ghane, “RECreate: A Blockchain-Based Marketplace for Renewable Energy Credits with Intelligent Peer-to-Peer Trading Using Model-Based Multi-Agent Reinforcement Learning”, IJRESM, vol. 9, no. 5, pp. 38–51, May 2026, Accessed: May 08, 2026. [Online]. Available: https://journal.ijresm.com/index.php/ijresm/article/view/3446

Most read articles by the same author(s)