Observer-based robust gain-scheduled control for semi-active air suspension systems subject to uncertainties and external disturbance

• Published in: Mechanical systems and signal processing Vol. 173; p. 109045
• Main Authors: Wang, Hang, Kin Wong, Pak, Zhao, Jing, Yang, Zhao-Xu, Yang, Zhi-Xin
• Format: Journal Article
• Language: English
• Published: Berlin Elsevier Ltd 01.07.2022; Elsevier BV
• Subjects: Active control; Actuators; Aerothermodynamics; Air springs; Control theory; Dynamical systems; Eigenvalues; Feedback control; Gain-scheduled control; Kalman filter; Kalman filters; Linear matrix inequalities; Linear parameter varying system; Nonlinear dynamics; Robust control; Semi-active air suspension system; Semiactive suspensions; Stiffness; Suspension systems; Transient response; Uncertainty; Linear parameter varying system; Gain-scheduled control; Semi-active air suspension system; Kalman filter
• ISSN: 0888-3270; 1096-1216
• DOI: 10.1016/j.ymssp.2022.109045

In this paper, an observer-based robust gain-scheduled control scheme is designed for a semi-active air suspension (SAAS) system subject to uncertainties and external disturbance. Since the air spring in SAAS system is a highly nonlinear component and its stiffness relies on the spring interior pressure and displacement, an air spring model is constructed based on aerothermodynamics theory. Aiming at comprehensive and simplified expression of air spring nonlinear dynamics, the air spring model is expressed as linear-parameter-varying (LPV) form and further integrated into a quarter car semi-active air suspension model. To identify system states under the noisy condition with lower cost, a LPV-Kalman filter is proposed to estimate the states which are difficult to be measured or physically unmeasurable in the semi-active air suspension. With the proposed observer, a robust gain-scheduled controller is developed to overcome complex nonlinear dynamics, actuator uncertainties and external disturbance. To realize a better transient response, the control law is improved by constraining eigenvalues of the closed-loop system in desired linear matrix inequality (LMI) region. At last, simulations and experiments are conducted to illustrate the effectiveness of proposed control scheme in a quarter car test rig.