Effects of Roll Dynamics in Model Predictive Control for Autonomous Vehicles

A Model Predictive Control (MPC) approach for autonomous vehicles is presented. We formulate a predictive control problem in order to best follow a given path by controlling the front steering angle. We start from the results presented in [4] and [7], where the MPC problem formulation relies on a simple bicycle model, and reformulate the problem by using a more complex vehicle model including roll dynamics. We present and discuss simulations of a vehicle performing high speed double lane change maneuvers where roll dynamics become relevant. The results demonstrate that the proposed model based design is able to effectively stabilize the vehicle by using a three dimensional vehicle model at the cost of a higher computational load