Stochastic model predictive controller with chance constraints for comfortable and safe driving behavior of autonomous vehicles

In this paper, we address the application of stochastic model predictive control with chance constraints to autonomous driving. We use a condensed formulation of a linearized vehicle model to setup a quadratic program with nonlinear chance constraints, which can be solved with off-the-shelf optimization algorithms. We further show how obstacle information in the path planning stage can be converted into a set of linear state constraints that can be directly used in the control algorithm. The resulting controller is potentially real-time capable and achieves a tradeoff between safety and comfort in its control behavior.