Abstract:
Numerous uncertainty factors in vehicle systems adversely affect the performance of the vehicle steering controller. To rectify this problem, we present a number of counter measures. First, a two-degree-of-freedom vehicle model is proposed; this model is based on the traditional bicycle model and considers possible unmodeled dynamics resulting from simplification of the bicycle model. Then, a steering controller that uses the stochastic model predictive control (SMPC) algorithm is designed to track lateral trajectory. To verify the effectiveness of the proposed algorithm, we perform simulations under various vehicle running conditions using the vehicle dynamics software (veDYNA). The simulation results show that the controller can achieve perfect tracking. To further verify the effects of considering unmodeled dynamics in the model, two controllers are designed and a series of experiments are conducted. The results suggest that the compensation for nonlinear motion can improve the accuracy of trajectory tracking.