Steering authority allocation strategy for human-machine shared control based on driver take-over feasibility
Chengrui Su,Haohan Yang,Jie Li,Xiaodong Wu
DOI: https://doi.org/10.1016/j.compeleceng.2024.109753
IF: 4.152
2024-10-17
Computers & Electrical Engineering
Abstract:Misoperation when a driver is distracted, e.g., turning the steering wheel unconsciously, preventing the machine from turning the steering wheel, etc., can lead to a false take-over of the vehicle, which poses a danger to shared control. Therefore, it is essential to evaluate the feasibility of driver take-over in distraction status. This paper proposes a steering authority allocation strategy based on driver take-over feasibility to reduce the interference of distracted driver behavior on shared control. First, a driver take-over feasibility framework is established based on quantified driver distraction levels and take-over intention. Then the shared control system is established based on a linear parameter-varying vehicle model and non-cooperative game theory to suppress the interference of time-varying speed and parameter uncertainty on shared control. Subsequently, the steering authority allocation strategy is constructed in conjunction with driver take-over feasibility and environmental risk evaluation. The proposed sigmoid-based mapping function ensures fast and smooth authority allocation. Finally, a series of driver-in-the-loop experiments are conducted to validate the effectiveness of our proposed strategy across various driving scenarios, and the results demonstrate that our strategy has superior performance than other advanced shared control schemes in reducing steering burden, improving driving safety, and enhancing vehicle stability, etc.
engineering, electrical & electronic,computer science, interdisciplinary applications, hardware & architecture