Optimizing safety and comfort for distracted drivers via adaptive shared control

Driver distraction introduces considerable uncertainty in vehicle operation, presenting significant challenges to driving safety and complicating human-machine collaborative control. To address these challenges, this paper proposes a shared control approach to support distracted drivers in maneuvering along extended curves. A model of distracted driver behavior is first constructed to simulate both path-following capability and distraction duration, capturing the operational characteristics of distracted drivers. Building upon this model, a multi-constraint model predictive controller (H-ORPF-MPC) is developed, integrating driver behavior and environmental factors to balance driving comfort and controller stability. Surrounding risks are quantified through a driver intention function and an Obstacle-Road Potential Field (ORPF), with real-time state constraint envelopes applied to control outputs, thereby ensuring vehicle operation within a safe range. Furthermore, safety and comfort-based authority allocation rules are formulated, accommodating both controller performance and driver preferences. Simulation results indicate that the proposed approach effectively assists distracted drivers, significantly reduces trajectory tracking errors, and improves driving comfort.