Human-Oriented Online Driving Authority Optimization for Driver-Automation Shared Steering Control

Abstract

Drivers are being confronted by the task of controlling vehicles together with automation systems before autonomous vehicles are even fully realized. A novel human-oriented online driving authority optimization shared steering framework is proposed to solve the problem of allocating driving authority between automation and human driver, in which the model predictive control(MPC) method is used to online optimize the driving authority. The merits of this framework is that, on the one hand, human driver workload can be alleviated in the case that the driver and the automation have similar driving intentions. On the other hand, the human driver has absolute control authority over an intelligent vehicle when the automation intention is inconsistent with that of the human driver. To illustrate the characteristics of the designed online driving authority optimization framework, simulations are carried out in the case of driver-automation cooperation and driver-automation non-cooperation. The results show that the proposed online driving authority optimization shared steering framework can not only restrict vehicles to safe areas but also lessen the driver’s workload.