Predictive modeling and active control of rollover in heavy vehicles

A 14 degrees-of-freedom vehicle model, a rollover prediction algorithm and an active rollover control algorithm for heavy vehicles are proposed in this study. The vehicle model is developed using Lagrange's equations as well as Newtonian mechanics, and used to support the development of the rollover prediction and active control algorithms. The developed vehicle model is verified by comparing its steady-state and transient response characteristics with those of ArcSim, which is significantly more complex than the proposed model. It has been shown that the responses of the proposed model closely match those of ArcSim under the same vehicle configuration and driving conditions.; The basic concept of the rollover prediction algorithm in this study is to predict vehicle rollover threat sufficiently in advance of rollover by comparing the predicted roll states with the corresponding rollover thresholds. To this end, this algorithm consists of two sub-algorithms: one to predict vehicle roll states using system identification and the other to identify rollover thresholds. These sub-algorithms are validated via simulation as well as via the test data of an actual vehicle. It has been shown that the roll states predicted using system identification are close to the simulated ones, and that the identified rollover thresholds are close to those obtained via simulation as well as via the tilt-table test of a tractor-semitrailer.; Two alternative control strategies: a speed and yaw-rate control system as well as a speed-only control system are developed for the active rollover control algorithm in this study. These control strategies, based on the concept of sliding control, prevent rollover by reducing the vehicle speed before cornering, and in the case of speed and yaw-rate control, reduce the impact of undesirable yaw response that may be generated during braking. The designed control systems are validated via simulation in presence of irregular road surface conditions. It has been shown that these control systems show good speed-tracking performance, and that the speed and yaw-rate control system shows better yaw responses than the speed-only control system.