Reference-free approach for mitigating human-machine conflicts in shared control of automated vehicles

Shared control is a promising approach that can facilitate the mutual understanding and cooperative control between human and machine. In this study, a novel reference-free framework for shared control of automated vehicles is proposed with the aim of mitigating conflicts between the human driver and the automatic system during their interactions. Position constraint and time to collision are deployed to prevent collision and guarantee stability by limiting the yaw rate and sideslip angle of the vehicle. The design of the shared controller is formulated as a model predictive control problem. It determines vehicle state based on the current driver command and implements control actions only if the vehicle state induced by the human driver violates the pre-defined constraints. The system models, safety constraints and the proposed shared controller are then integrated. The algorithm is validated through computer simulation under two different driving scenarios. The simulation results show that the proposed reference-free approach can offer the human driver more degrees of freedom during shared control, effectively mitigating human-machine conflicts, compared to the previous work. In addition, the algorithm ensures the safety and stability of the system under risky driving conditions, validating its feasibility and effectiveness.