Dynamics modeling and nonlinear observer design for vehicle rollover avoidance

A real-time accurate detecting the risk of an impending vehicle rollover is essential for active rollover avoidance systems. To this end, a 4DOF nonlinear vehicle roll dynamics model is developed, which fully exploit the coupling effects between the yaw dynamics, lateral dynamics, and roll dynamics. An asymptotically stable observer is presented by using the mean value theorem of vectorial form, and the observer gain is derived based on linear matrix inequalities (LMIs). The simulations demonstrate that the developed nonlinear observer can provide reliable estimation of the vehicle states, even under aggressive driving conditions.