Model-based safe coordinated control for centrifugal compressors using evolution strategy
Results in Engineering, vol. 30, pp. 111297
Abstract
Centrifugal compressors are essential components in fluid machinery systems, often operating under variable load conditions that demand rapid and safe adjustment of control parameters. However, their highly nonlinear and strongly coupled nature introduces significant challenges in maintaining system stability, especially under transient conditions where overshoot can lead to surge or even failure. To address these challenges, this paper proposes a coordinated control strategy that integrates a lumped-parameter model for real-time surge risk assessment with a Covariance Matrix Adaptation Evolution Strategy (CMA-ES) for online optimization of PID controller parameters. The model quantifies the safety margin during setpoint transitions, while the CMA-ES optimizer simultaneously tunes multiple PID gains without requiring system decoupling. The proposed strategy is validated through simulations under three industrially relevant operating modes: constant outlet pressure control, constant mass flow control, and coordinated pressure-flow control. Results demonstrate that the method significantly improves safety and robustness compared to conventional Ziegler-Nichols tuning, effectively suppressing flow and pressure overshoot while maintaining comparable adjustment speed. These findings highlight the strong potential of the proposed approach for practical engineering applications.
Authors 6
-
Leiden University · Xi'an Jiaotong University
Affiliation as printed
Leiden Institute of Advanced Computer Science, Leiden University, Leiden 2333 CA, Netherlands
School of Mechanical Engineering, Xi′an Jiaotong University, Xi′an 710049, China
-
Qing Zhang corresponding
Affiliation as printed
School of Instrument Science and Technology, Xi'an Jiaotong University, Xi'an 710049, China
-
Affiliation as printed
Xi'an Huayun Wisdom Information Technology Co., Ltd, Xi’an, 710075, China
-
Affiliation as printed
School of Energy and Power Engineering, Xi'an Jiaotong University, Xi’an, 710049, China
-
Affiliation as printed
Leiden Institute of Advanced Computer Science, Leiden University, Leiden 2333 CA, Netherlands
-
Affiliation as printed
Leiden Institute of Advanced Computer Science, Leiden University, Leiden 2333 CA, Netherlands
Cited by 0 stored of 0
No patents citing this paper on Lens.org (checked 2026-10-11).