Automobile drivers appreciate low interior noise levels as a positive quality feature. The automatic evaluation of engine and gearbox noise levels as well as production fault detection are fixed parts of car quality tests. Typical tasks which have to be performed in this context are sensor choice and attachment, data acquisition, preprocessing, computation of psychoacoustic and statistical features and decision making. Artificial neural networks and fuzzy logic are adaptive pattern recognition and decision making approaches, which are currently under investigation. The aim of this contribution is the efficiency analysis of combined concepts using neural networks and fuzzy logic for the determination of gear noise levels and failure detection. Signals from an experimental gear and from passenger car rear axle gears serve as input data. A DSP supported parallel processor system developed in cooperation with a major automobile manufacturer provides the necessary computational power.
Acoustic quality test and fault detection of automobile gearboxes using artificial neural networks and fuzzy logic
Akustische Qualitätsprüfung und Fehlererkennung von Kraftfahrzeugwechselgetrieben mit Hilfe von künstlichen neuronalen Netzen und Fuzzy-Logik
1995
8 Seiten, 6 Bilder, 4 Tabellen, 11 Quellen
Conference paper
English
Automotive engineering | 1995
|British Library Conference Proceedings | 1995
|Fuzzy connectionist network for fault diagnosis of helicopter gearboxes
Tema Archive | 1995
|Automotive engineering | 1993
Fault Classification by Neural Networks and Fuzzy Logic
British Library Conference Proceedings | 1995
|