Rollover Prediction and Control Strategy Based on Experiment for Tractor Semitrailer Fitted with Hydraulically Interconnected Suspension

Hydraulically interconnected suspension (HIS) system has theoretically and experimentally been proved to suppress rollover of non-articulated vehicle, but lacking of vibration studies on more complex nonlinear systems such as multi-axle combination vehicle. Hence, this paper presents rollover threshold prediction and active disturbance rejection controller (ADRC) of a tractor semitrailer fitted with HIS based on experimental test to reduce instability of mechanical system. Firstly, a roll resistant verification of a tractor semitrailer fitted with HIS is conducted under a steady circle experiment. Furthermore, the prediction for rollover threshold to extreme maneuver is presented by incorporating robust Kalman filter (RKF) and Back Propagation neural network with genetic algorithm (GA-BP) that are both designed. Finally, an ADRC control is proposed based on control model reduced by applying least square identification and model reduction method. The results are obtained including three aspects: experimental results verify the superior anti-roll performance of vehicle fitted with HIS in comparison with traditional vehicle and are also used to assist the subsequent prediction at high speed. Prediction results highlight that the RKF algorithm performs better filtering accuracy over KF, UKF and PF methods and developed GA-BP supply the higher precision compared with the conventional BP method for the prediction, by which the lateral acceleration threshold is achieved. The results of control show ADRC control strategy obviously has faster response and smaller overshoot than fuzzy PID and regular PID controller for the stability of system.