Deep Reinforcement Learning Home Energy Management Based Local Energy Trading in a Blockchain Platform

Abstract

Decentralized peer-to-peer Local Energy Markets (LEMs) are gaining popularity as local power production from Renewable Energy Sources (RESs) increases. The study investigates a blockchain-driven LEM in which prosumers and consumers buy and sell energy independently of the involvement of a third party. The suggested scheme involves Home Energy Management (HEM) and demurrage mechanisms that enable both consumers and prosumers to reduce their power prices and improve their power usage. The end-user can likewise move the load to off-peak periods and benefit from lower energy costs from the LEM through the approach. In this solution, HEM and demurrage mechanisms are used to optimize power usage as well as energy costs. As well as providing adequate power for the LEM, it is economically beneficial for end users and the community. The proposed system in the study utilizes the Deep Deterministic Policy Gradient (DDPG) algorithm and demurrage mechanisms to optimize electricity usage and energy costs. Meanwhile, smart contract on the Ethereum blockchain regulate and safeguard the electricity trading process.