ADAS Fail-Safe를 위한 센서 검출 오류 발생시 전방 차량 트랙 관리 시스템에 관한 연구

Abstract

Adavanced Driver Assistance System(ADAS) use a variety of sensors as their functions become more complex. Errors from each sensor cause fatal malfunctions in the entire system. Therefore a multi-sensor Fail-Safe algorithm for Advanced Driver Assistance System(ADAS) has been considered in this paper. We proposed a sensor track management system which diagnoses failure time of either Radar or Camera that lead to a fatal malfunction under the ADAS application performing. The system operates multi-sensor fusion with forward radars and a monocular camera under the supervision of Track Management System. The main purpose in this research lies in the field of track loss from vehicle detection fault. We have tested several simulation tests for analysis on the vehicle detection fault of time for ADAS Fail-Safe and three important tests of the results demonstrated that perception module among the entire ADAS system has an important role and should keep on diagnosing the sensor fail to hand it over to the next decision and control.