Blockchain and Quantum Machine Learning Driven Energy Trading for Electric Vehicles

• Published in: Ad hoc networks Vol. 165; p. 103632
• Main Authors: Kashyap, Pankaj Kumar, Dohare, Upasana, Kumar, Manoj, Kumar, Sushil
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.12.2024
• Subjects: Blockchain; Electric Vehicles; Energy Trading; Quantum Reinforcement Learning; Smart Contract; Smart Contract; Blockchain; Electric Vehicles; Quantum Reinforcement Learning; Energy Trading
• ISSN: 1570-8705
• DOI: 10.1016/j.adhoc.2024.103632

With the steep growth of Electric Vehicles (EV's), the consequent demand of energy for charging puts significant load to powergrids. Renewable Energy Sources enabled microgrids can alleviate the problem of energy demand and trade the energy locally in Peer-to-Peer (P2P) manner, where seller (microgrid) and buyer (EV's) “meet” to trade electricity directly on agreed term without any intermediary. However, a foolproof system required for audit and verification of transaction record between seller and buyer to address privacy and security in untrusted and opaque local energy trading market (LETM). Centralized public blockchain enabled system (for audit the transaction records and storage) based on conventional learning models faces mainly two issues in the LETM. (a) if, centralize system runs out of energy and tear down then whole energy trading plunges treated as single point of failure (b) Conventional learning models fail to converge optimal point in case of large state and action space (large number of EV's and their energy demand). The primary objective of this paper to provide secure system for LETM, 1) Distributed nature of Consortium Blockchain used that solve the problem of single point of failure to audit and storage of transaction and profile info of microgrids and EV's. 2) Quantum based Reinforcement Learning (QRL) easily handles the large number of EV's energy supply and demand for smoothly run LETM. In this context, this paper presents Blockchain and Quantum Machine Learning driven energy trading model for EVs (B-MET). A utility maximization problem formulated as Markov Decision Process (MDP) and their solution provided by using QRL focusing on join optimization of selling price, loan amount and quantity of shared energy. MDP is a mathematical framework used to model decision-making in situations where outcomes are partly random and partly under the control of a decision-maker, i.e., the future state depends only on the current state and action, not on the sequence of events that preceded it. QRL method combines quantum theory with traditional RL. It is inspire by the principles of state superposition and quantum parallelism. Convergence analysis and performance results attest that B-MET convergences faster, maximizes the utility with lower confirmation delay in P2P energy trading as compare to state of the art techniques.