Research on robust backstepping sliding mode control strategy for magnetorheological semi-active suspension based on quantized input signals

Abstract

To enhance control effectiveness and mitigate system oscillations in backstepping sliding mode control, a robust backstepping sliding mode control strategy based on a hysteretic quantizer is proposed for application in magnetorheological (MR) semi-active suspension (SAS) systems. To accurately predict the output current for precise control, a fuzzy neural network (FNN) is used to construct an inverse model for magnetorheological damper (MRD), input the current into the Dahl model and output the actual damping force. Based on the MRD inverse model, the robust term is introduced in the backstepping sliding mode control, and a hysteretic quantizer is used to quantify the controller input signals, avoiding the occurrence of oscillations and forming a complete closed-loop feedback control. Simulation results demonstrate that the FNN inverse model can accurately predict the control current. Compared to backstepping control and passive suspension, the robust backstepping sliding mode control strategy based on a hysteretic quantizer (RBSMC-HQ) reduces the root mean square (RMS) values of vehicle body acceleration by 37.59% and 44.82% respectively under A-Class road conditions, and by 41.93% and 44.84% respectively under B-Class road conditions. This control strategy effectively enhances the ride comfort and handling stability.