作者: Mariela Cerrada , Chuan Li , René-Vinicio Sánchez , Fannia Pacheco , Diego Cabrera
DOI: 10.1016/J.FSS.2016.12.017
关键词: Machine learning 、 Fault severity 、 Fault (power engineering) 、 Path (graph theory) 、 Hierarchical clustering 、 Mathematics 、 Fuzzy logic 、 Artificial intelligence 、 Failure mode and effects analysis 、 Data mining 、 Vibration 、 Field (computer science)
摘要: Abstract Rotating machinery is an important device supporting manufacturing processes, and a wide research works are devoted to detecting diagnosing faults in such machinery. Recently, prognosis health management rotating have received high attention as area, some advances this field focused on fault severity assessment its prediction. This paper applies fuzzy transition based model for predicting conditions helical gears. The approach combines Mamdani models hierarchical clustering estimate the membership degrees levels of samples extracted from historical vibration signals. These used weighted transitions modelling evolution along states over time, according certain degradation path. obtained able one step-ahead failure mode under study, by using current previous two available successive input samples. predictive was validated real data test bed with different damages tooth breaking Results show adequate predictions scenarios paths.