Driver Speed Control Modeling for Predictive Braking Assistance System Based on Risk Potential in Intersections

Abstract

This paper describes the assessment of a predictive braking assistance system, which is done using a driving simulator that reconstructs near-miss incident scenarios relevant to pedestrians. An autonomous braking assistance algorithm for collision avoidance is designed based on pedestrian movement prediction and an artificial risk potential field. A virtual spring connecting the vehicle and the pedestrian is used to determine the repulsive potential field and the intensity of the deceleration. The feasibility of the proposed braking assistance algorithm is examined through experiments using the driving simulator and a comparison to actual driving data. Near-miss incident data relevant to pedestrians in intersections are analyzed to get the basic parameters of a crash scenario model relevant to pedestrians. Driving simulator experiments are used to verify the effectiveness of the proposed system.