Integrated Obstacle Detection and Avoidance in Motion Planning and Predictive Control of Autonomous Vehicles
Rien Quirynen, Karl Berntorp, Karthik Kambam, Stefano Di Cairano
Abstract
This paper presents a novel approach for obstacle avoidance in autonomous driving systems, based on a hierarchical software architecture that involves both a low- rate, long-term motion planning algorithm and a high-rate, highly reactive predictive controller. More specifically, an integrated framework of a particle-filter based motion planner is proposed in combination with a trajectory-tracking algorithm using nonlinear model predictive control (NMPC). The motion planner computes a reference trajectory to be tracked, and its corresponding covariance is used for automatically tuning the time-varying tracking cost in the NMPC problem formulation. Preliminary experimental results, based on a test platform of small-scale autonomous vehicles, illustrate that the proposed approach can enable safe obstacle avoidance and reliable driving behavior in relatively complex scenarios.