This paper presents an obstacle avoidance scheme for autonomous vehicles as an active safety procedure in unknown environments. The obstacle avoidance problem is treated using a nonlinear model predictive framework in which simplified dynamics are used to predict the state of the actual vehicle over the look-ahead horizon. Due to the slight dissimilarity between the simplified model used for trajectory generation and the actual vehicle trajectory, a separate tracking controller is designed to track the generated trajectory. The longitudinal dynamics of the vehicle is controlled using the inverse dynamics of the vehicle power-train model and the lateral controller is designed based on the linear quadratic regulator. In the nonlinear model predictive framework, the threat of local obstacles is augmented into the performance index using a parallax-based method. The simulation results show that the presented model-predictive-control-based trajectory generation and tracking controller, together, give satisfactory performance in terms of obstacle avoidance when applied to the full nonlinear vehicle model.
展开▼