Online Robust MPC based Emergency Maneuvering System for Autonomous Vehicles

TL;DR

This paper presents a real-time robust tube MPC framework for emergency steering in autonomous vehicles, integrating path planning and following to improve safety during obstacle avoidance.

Contribution

It introduces a novel real-time capable robust tube MPC framework that combines path planning and following for emergency maneuvers in autonomous vehicles.

Findings

Enhanced robustness over non-robust MPC in simulations

Effective handling of extreme maneuvering scenarios

Improved safety during emergency obstacle avoidance

Abstract

Nonlinear Robust Model Predictive Control (RMPC) provides a very promising solution to the problem of automatic emergency maneuvering, which is capable of handling multiple possibly conflicting objectives of robustness and performance. Even though RMPC gives a suboptimal solution, the key challenge in real-time implementation is that it is computationally very demanding. In this paper a real-time capable robust tube MPC based framework for steering control during emergency obstacle avoidance maneuver is presented. The novelty of this framework lies in the robust integration of path planning and path following tasks of autonomous vehicles. A simulation study showcases the robust performance improvements due to the proposed strategy over a non-robust MPC in different extreme maneuvering scenarios.