Proximal Policy Optimization Algorithm for Integrated Energy System Operation with Adaptive Learning Rate Decay Strategy

Authors:

Abstract
With increasing focus on sustainability and efficiency, Integrated Energy Systems (IES) have gained more attention in the provision of electricity and thermal energy. However, the inherent complexity and uncertainty of IES pose challenges to their optimization and management. This paper proposes a Proximal Policy Optimization (PPO) algorithm for the operation of IES, based on an adaptive learning rate decay strategy, aimed at enhancing the operational efficiency and stability of IES. Initially, a Markov decision process is established to simulate the operation of IES. Subsequently, a PPO algorithm with an adaptive learning rate decay strategy is proposed. Experimental results demonstrate that the proposed algorithm significantly outperforms traditional PPO algorithms in reducing overall operational costs and improving the balance of electricity and thermal energy supply and demand. This research provides an effective optimization method for the efficient and sustainable operation of IES.
Funding Information
  • Research and Development

This publication has 10 references indexed in Scilit: