Hybrid architecture for vehicle lateral collision avoidance

Abstract

In this study, the authors consider the problem of collision avoidance of ground vehicles. In this framework, they focus on a single vehicle, and they guarantee that a given safe set is not exited during a mission. More specifically, they consider the yaw-lateral motion of a two-wheel vehicle subject to a side wind, and they design a state-feedback controller such that the vehicle follows the nominal trajectory with a given maximum error. For these purposes, they resort to a control system where a supervisor, implemented as a hybrid automata, decides to switch between two controllers, depending on the measured vehicle lateral displacement from the nominal trajectory. The two controllers are designed using an H ∞ approach and the input–output finite-time stability approach, respectively. Some simulation results are included, showing the effectiveness of the proposed architecture.