A Stereo Vision-Aided Receding Horizon Vehicle Dynamic Control System for Independent Drive Vehicles Steered by Wire

Abstract

This paper proposes a stereo vision-aided receding horizon vehicle dynamic control system for independent drive vehicles steered by wire. First, we propose the receding horizon vehicle dynamic control (RHVDC) to track the desired slip angle and yaw rate of a dynamic vehicle. The RHVDC computes and updates control inputs at each sampling time to minimize a quadratic cost function over the finite future time horizon, allowing it to address the time-varying property of dynamic vehicles. Its stability is shown to be guaranteed with appropriate parameters. Second, unlike the conventional steering systems fixed with the driver's steering wheels, the SBW system improves the controllability of vehicle dynamics and enables better tracking performance. Third, combining a stereo vision system with existing sensors allows state estimation of dynamic vehicles to be more accurate in noisy environments. The three strategies mentioned above ultimately maximize the tracking performance of dynamic vehicles through improvement in terms of theory, facilities, and recognition. Through the software-in-the-loop simulation (SILS) with CarSim, the proposed RHVDC system has the better tracking performance for controlling both the slip angle and the yaw rate compared with the existing ones. It is also observed that the performance enhancement of the proposed RHVDC system is more outstanding on slippery or more severe roads.