Robust Lane Keeping Control in Automated Vehicles: A Driver-in-the Loop Approach

The shared control between human driver and autonomous controller for vehicles has been evaluated over several driving conditions such as cruise control, lane keeping/ lane change assist, highway driving etc. Typically, these approaches synthesize controllers based on linear vehicle models integrated with visual cues based driver model. In this work, to consider the aspect of tire-friction non-linearity, Brush-Tire force model has been considered. Further, to address conflict between human driver and autonomous controller during typical maneuvers, a novel sharing parameter was introduced in the proposed robust sliding mode based control design. Subsequently, the proposed shared control technique is validated thru numerical simulations over the Satory test track, with lateral wind force and road curvature considered as disturbance. Discussions on influence of varying road-friction conditions and sharing parameter proposed is also presented.