A concept of fault detection and diagnosis is presented which is applied to fast and closed-loop complex systems. In the future, the aim of this application will be to improve road safety, by controlling actuators when a critical situation is detected. For the present, it is only to provide information about the criticality of the situation. As a real-time process was necessary, the difficulty had lain in conceiving a time-saving algorithm (the frequency of data acquisition was 50 Hz). An other necessity was to provide a control diagnosis properly correlated with human expert opinion. Five critical driving situations are dealt with: foot off the accelerator, banking, sidewind gust, asymmetrical braking and braking while cornering. The supervision structure is described, which is based upon a sequential classification of detected faults and upon the use of the residues of two models. The diagnosis itself is built with a fuzzy logic algorithm. The data were provided by several real car tests. The system is currently being implemented and tested in real road conditions.
Real time fault detection and diagnosis: Application to critical behaviour of a road vehicle
Echtzeitfehlererkennung und Fehlerdiagnose für Straßenfahrzeuge in kritischen Situationen
1996
9 Seiten, 11 Bilder, 9 Quellen
Aufsatz (Konferenz)
Englisch
Real time fault detection and diagnosis: application to critical behaviour of road vehicle
Kraftfahrwesen | 1996
|Real Time Fault Detection and Diagnosis: application to Critical Behaviour of a Road Vehicle
British Library Conference Proceedings | 1996
|Vehicle-mounted real-time fault diagnosis method and device
Europäisches Patentamt | 2021
|Agent-based Real-time Fault Diagnosis
British Library Conference Proceedings | 2005
|Robust Real-time Lane and Road Detection in Critical Shadow Conditions
British Library Conference Proceedings | 1995
|