Real-time roll state estimation and rollover prediction for light SUVs

This paper presents a roll state estimator and a rollover prediction model based on a real-time roll estimator. The real-time roll estimator is designed to estimate the current roll angle and roll rate of the vehicle body from a 3DOF vehicle model. The veDYNA dynamic simulation software was used to verify the performance of roll motion prediction. The prediction model calculates the time it takes for the inside tire lift-off to determine the danger of rollover based on current roll angle and roll rate. Obstacle avoidance and ramp entering maneuvers are designed to verify the prediction model. Simulation results with driver model in the loop show that the prediction model could represent the severity of the rollover threat accurately across all the driving scenarios.