An Improved Vehicle Rollover Prediction Algorithm

The extended 3-DOF (degree-of-freedom) nonlinear rollover prediction model for heavy-duty vehicles is presented by considering the nonlinear characteristic of tire. The lateral force of rear wheels is revised with lateral acceleration of center of gravity, aiming at reflecting the impact of vehicle cornering condition. The parameters of current vehicle status are obtained in real time by using extended Kalman filter technique, and the vehicle roll state is estimated based on lateral load transfer ratio (LTR) algorithm. Simulation results show that the extended model is correct and more effective in predicting the current vehicle status and rollover tendency. In addition, the prediction results of our model are much closer to the practical data obtained in vehicle tests.