Parameter Effects of the Potential-Field-Driven Model Predictive Controller for Shared Control

Abstract

Parameter effects of the potential-field-driven model predictive control (PF-MPC) method on performances of shared control systems during obstacles avoidance are investigated. The PF-MPC controllers of autonomous driving and shared control systems are designed based on the constructed potential fields and model predictive control method, and the driver-vehicle dynamics and the driver-related costs are also considered in the design of the shared controller. To explore a potential approach of alleviating driver-automation conflicts of the shared control systems, different motion planning results generated by the PF-MPC controller are explored by adjusting effects of potential fields’ parameters, which provides possibilities to decrease driver-automation conflicts between the planned trajectory and driver’s target path. Moreover, two case studies are designed to discuss different frameworks and parameters effects on shared control systems. Results show that the proposed shared control frameworks considering driver-vehicle dynamics and the driver-related cost show better performances regarding driver-automation conflicts management and driving safety than the decentralized control framework. And the longitudinal normalized constant of potential fields parameters shows influences on the driver-automation conflicts management and driving safety performances of shared control.