Rollover Index Estimation in the Presence of Sensor Faults, Unknown Inputs, and Uncertainties

Abstract

The rollover status of a vehicle indicated by the lateral load transfer (LTR) as the vehicle traverses over various driving scenarios is critical in the implementation of antirollover control procedures. The determination of LTR is often carried out by the measurements of the roll angle, the lateral acceleration, vertically acting suspension forces, etc. In all these measurements, sensor faults may occur, which lead to a faulty computation of the rollover status, raising a false alarm. In this work, a scheme based on a robust higher order sliding-mode observer is proposed to estimate the states of a nonlinear two-wheel vehicular system affected by road disturbances, uncertainties in the height of the vehicle's center of gravity, and possible multiple sensor faults. Applying the proposed approach, the unknown inputs and sensor faults are reconstructed. To perform the estimations, adaptive sliding-mode-based observers that do not require the knowledge of the bounds of the uncertainties and unknown inputs are designed. Consequently, the true rollover status of the vehicle is determined, in spite of the presence of sensor faults. The validity of the proposed scheme has been assessed on the vehicle simulation software CarSim as the vehicle undergoes a double lane change maneuver.