Research on Surge Control of Centrifugal Compressor Based on Reinforcement Learning

Kun Jiang,Yang Xiang,Tianyou Chen,Chaojun Jiang

DOI: https://doi.org/10.1007/978-3-030-57745-2_25

2020-01-01

Abstract:Surge control is always an important issue in compressor health management. Although the traditional PID anti-surge methods has made some achievements, the adaptation across variable working conditions is still a challenge. Meantime, reinforcement learning (RL) has been proven strong adaptability and high accuracy in recent years, however, few researches attempt RL in surge control applications. Therefore, this paper proposes a RL anti-surge method for compressor. In specific, a novel architecture based on Actor-Critic algorithm is presented, and the relative structures are re-designed aiming at the issue. Moreover, the simulation of supercharging system is conducted to validate the proposed method, in which the compressor operation models are static and dynamic, respectively. The results show that the agent of Actor-Critic can sense the state of the compressor in real time and then output the appropriate anti-surge valve opening which not only avoids the surge, but also ensures the working efficiency at a high level. Comparing with the traditional PID anti-surge method, the superiority of the proposed method such as the strong adaptability across variable working conditions is verified.