Abstract:

With the development of the EV technology， vehicle active suspension systems with onboard sensors have become increasingly widespread. However， due to the existence of measurement noise in sensors， classic control algorithms struggle to gain full-state observation under mismatched disturbances of non-Gaussian measurement noise， which severely affecting the performance of the suspension control systems. In this paper， for the challenges from mismatched disturbances with measurement noise in active suspension sensors， as well as the matched disturbances of the suspension intrinsic nonlinear and unmodeled dynamics， a high-order sliding mode observer （SMO） is designed to gain full-state observation for both the sprung and upsprung subsystems. Additionally， a fixed-time sliding mode controller （FSMC） is employed to establish a globally convergent closed-loop control system for the active suspension system. The simulation results show that the proposed SMO-FSMC algorithm significantly improves ride comfort compared to LQG-based and extended observer-sliding mode control algorithm under random road profile. Furthermore， the proposed system exhibits superior state tracking accuracy for the active suspension and greater robustness against external disturbances and measurement noise compared to Kalman filters and extended observers. Besides， under speed bump scenarios， the SMO-FSMC system can effectively attenuate the vertical acceleration of the sprung mass thus improving the ride comfort.

Key words: active suspension, mismatched disturbance, fixed-time sliding mode control, higher order sliding mode observer