Real-time rollover prediction for vehicle based on principles of sliding mode and fuzzy inference system

This paper presented a real-time rollover prediction algorithm based on roll state estimator. An extended 3-DOF vehicle model considering tire nonlinear characteristics was firstly proposed, the parameters of the tire model were obtained by using nonlinear least square method. Vehicle roll state was estimated using sliding mode(SM) observer based on the super twisting algorithm. A rollover critical index(RCI), which indicated the risk of rollover, was developed by a fuzzy inference system(FIS) based on current vehicle state and its trend. The veDYNA dynamic simulation software was used to verify the performances of roll state estimator and roll motion prediction.