Trust-centric shared control with driver NMS characteristics for autonomous vehicles

To mitigate conflicts between human drivers and automated systems in autonomous vehicles, this paper proposes a trust-aware shared control strategy. A driver–automation interaction model incorporating driver neuromuscular dynamics is first established, with uncertain parameters identified via a hierarchical least squares method. Building upon this model, a trust-driven policy is constructed by quantifying driver trust and its influencing factors through a dynamic trust map. Furthermore, a meta-game-theoretic allocation mechanism is introduced to dynamically distribute control authority between the driver and automation system based on optimal utility. Finally, a hardware-in-the-loop platform integrating a driving simulator, electromyography sensors, and a dSPACE module is implemented for validation. Experimental results demonstrate that the proposed method effectively enhances driving stability, reduces driver–automation conflict, and improves trust-aware cooperative control performance.