Driver Assisted Lane Keeping with Conflict Management Using Robust Sliding Mode Controller.

Area of Science:

• Automotive Engineering
• Control Systems
• Human-Machine Interaction

Background:

• Lane-keeping assistance systems face challenges in balancing lane tracking, driver comfort, vehicle stability, and minimizing driver-controller conflict.
• Existing systems often struggle with seamless transitions between manual and autonomous control, leading to potential conflicts.

Purpose of the Study:

• To propose a cooperative control strategy for lane-keeping assistance that integrates driver monitoring, adaptive assistance allocation, and human-in-the-loop control.
• To develop a novel higher-order sliding mode controller for smooth authority transitions and establish closed-loop stability.
• To minimize driver-controller conflict using a new sharing parameter.

Main Methods:

• Identified a time-varying driver loading pattern based on lateral acceleration, road curvature, and driver torque.
• Formulated an adaptive driver activity function using monitored driver state and loading patterns.
• Developed a higher-order sliding mode controller for seamless mode transitions and introduced a novel sharing parameter to mitigate conflict.

Main Results:

• Experimental validation on a high-fidelity simulator demonstrated real-time implementation feasibility.
• Field tests showed a 9.4% improvement in cooperative driving quality.
• Significant reductions in steering workload (86.13%) and driver-controller conflict (65.38%) were achieved compared to existing designs.

Conclusions:

• The proposed cooperative control strategy effectively enhances lane-keeping assistance performance.
• The novel controller and sharing parameter significantly improve cooperative driving quality, reduce workload, and minimize conflict.
• This approach holds significant potential for addressing various road transportation challenges in autonomous driving.