Dynamic Pricing Integrated Demand Response for Multiple Energy Carriers with Deep Reinforcement Learning

Abstract

The traditional scope of demand response has been expanded to include integrated demand response (IDR), leveraging the technological sophistication provided by energy integration technologies. This paper examines the relationship between service providers (SP) and end-users in the IDR programme. The purpose of IDR is to maximise profits for gas and electricity utility companies while also minimising customer consumption prices and keeping the system stable. The hierarchical decisionmaking framework is illustrated using deep reinforcement learning (DRL). To address this challenge, the deep deterministic policy gradient (DDPG) technique uses deep neural networks to assess the state and compute the action. SP can adjust retail energy pricing adaptively during the online learning process, Considering end-user demand uncertainty and wholesale price flexibility. Experiments demonstrate that our proposed approach achieves high performance. The findings demonstrate that The IDR programme can benefit both the end-users and the provider by lowering energy costs and peak load demand.