Enhancing Controllability and Path Tracking with Overactuated Autonomous Vehicles

Abstract

Motion control of autonomous vehicles presents unique challenges in terms of stability and obstacle avoidance. Especially in critical driving conditions such as μ-low, autonomous vehicles may reach the limits of stability and require the use of all available lateral friction to maintain control and stably follow a reference path, while also taking into consideration collision avoidance. Overactuated vehicles can mitigate the issue of limited stability and controllability by optimal use of the lateral friction potential by using additional actuators. To address these challenges, this paper introduces a novel control architecture that combines a linear model predictive controller for tracking with model following and nonlinear control allocation technique. The control allocation method uses torque vectoring and rear wheel steering to extend the linear understeer characteristics and increase the maximum achievable lateral acceleration. Simulations demonstrate that this method reduces the path tracking error and enhances stable maneuvering in critical driving conditions by increasing the vehicle’s lateral capabilities.