Obstacles detection and collision avoidance system developed with virtual models

Abstract

Insurance companies have notice that since 1984, a couple of years after antilock braking systems (ABS) were introduced in market, traffic accidents and its injuries have been decreased. Developments in electronics and mechanics have improved the vehicle performance in collision. Lately several developments on preventing or avoiding collision have raised as active safety systems or so called in Europe, Advanced Driver Assistance Systems (ADAS). Systems as adaptive cruise control, lane change assist and blind spot detection have been developed facing the challenge of avoiding collision. In present paper is shown a model of collision avoidance for automotive applications. The system includes a model for vehicle dynamics: it was developed with the causal software AMESIM. Decision functions were developed to determine when an object is a dangerous obstacle, those functions depends on relative speed, and distance between host vehicle and obstacle. Vehicle model and decision functions are integrated to become a system for collision avoidance. The system warns the driver in a distance safe enough to avoid the collision in case the driver neglects warnings, the system begins braking in order to decrease damage severity if the collision happens or even avoid it. The simulation results of selected collision scenarios are presented. Also a brief description of available sensors for this application is shown.