Supervision, fault detection, and sensor fault tolerance of passenger cars

Abstract

In the not too distant future, driver assistance systems (like a collision avoidance system) need more and more accurate infonnation about the current vehicle driving situation. This information can be obtained by a driver situation supervision system. This supervision unit can then be used as driver warning system or for new vehicle control algorithms. After a short introduction to a lateral vehicle model, a deterministic approach for the online calculation of different driving conditions (i.e. stability, understeering, oversteering, and neutralsteering) is given. But, driver assistance systems also need reliable sensor infonnation for decision-making. A presented sensor fault detection and fuzzy logic fault diagnosis system can classify 14 different faults. the fault type, fault location and fault cause. In the case of a faulty sensor signal, a reconfiguration of the sensor system is done by a sensor fault tolerance system. Real test-drives were done to verify the results physically.