Abstract
This paper proposes an evaluation methodology of safe driving based on the driver cognitiveoperation model in crossing an intersection situation. To prevent the intersection accidents, we can employ two approaches: one is braking assistance which activates to stop the vehicle when there is a possibility of crash accident, the other is advisory and warning systems which inform drivers about dangerous situations and their present driving habits. Against this method, we propose an alternative system that evaluates the driving behavior at intersection from the viewpoint of cognitive behavior by using the head-yaw angle as an index of safety confirmation, and the viewpoint of operation behavior by focusing on braking behavior. The paper describes a micro electric vehicle which is used as an experimental vehicle and mounted with a driving data recording system including a DGPS sensor, and a 3-D driver-head motion sensor. In the first step, it addresses the deceleration behavior at signalized intersections of a number of drivers. The velocity at deceleration start and stopping distance of five drivers are analyzed. In the second step, the safety confirmation behavior based on driver head yaw motion and the vehicle velocity pattern during crossing a non-signalized intersection without right of way is analyzed.
Keywords: Automobile, Active safety, Intersection, Driver modeling, Driver assistance