Surrogate Empowered Sim2Real Transfer of Deep Reinforcement Learning for ORC Superheat Control

TL;DR

This paper introduces a Sim2Real transfer learning approach for deep reinforcement learning in ORC superheat control, enhancing training efficiency and generalization across varying operating conditions in energy systems.

Contribution

It presents a novel Sim2Real transfer learning method that improves DRL training speed and robustness for ORC control, addressing safety and generalization challenges.

Findings

Significantly faster DRL training in ORC control tasks.

Enhanced generalization of the DRL agent across multiple operating modes.

Effective mitigation of safety risks associated with physical system testing.

Abstract

The Organic Rankine Cycle (ORC) is widely used in industrial waste heat recovery due to its simple structure and easy maintenance. However, in the context of smart manufacturing in the process industry, traditional model-based optimization control methods are unable to adapt to the varying operating conditions of the ORC system or sudden changes in operating modes. Deep reinforcement learning (DRL) has significant advantages in situations with uncertainty as it directly achieves control objectives by interacting with the environment without requiring an explicit model of the controlled plant. Nevertheless, direct application of DRL to physical ORC systems presents unacceptable safety risks, and its generalization performance under model-plant mismatch is insufficient to support ORC control requirements. Therefore, this paper proposes a Sim2Real transfer learning-based DRL control method for ORC superheat control, which aims to provide a new simple, feasible, and user-friendly solution for energy system optimization control. Experimental results show that the proposed method greatly improves the training speed of DRL in ORC control problems and solves the generalization performance issue of the agent under multiple operating conditions through Sim2Real transfer.