Open AccessFault Prediction of Fan Gearbox Based on Deep Belief Network
Author(s) -
Hanfeng Zheng,
Yiru Dai
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1449/1/012050
Subject(s) - deep belief network , fault (geology) , computer science , artificial intelligence , artificial neural network , feature extraction , noise (video) , deep learning , feature (linguistics) , pattern recognition (psychology) , raw data , data mining , machine learning , linguistics , philosophy , seismology , image (mathematics) , programming language , geology
Due to the large amount of data in the actual fan equipment failure, the external noise is complicated, and there is a high degree of nonlinearity and complexity, which makes it difficult to extract the fault features. If the model is constructed by the traditional method, the accuracy of the fault prediction is poor. Therefore, considering the advantages of deep learning in data feature extraction, this paper proposes a wind fault prediction method based on deep belief network (DBN). The original raw data is firstly deleted and normalized, and then imported into the DBN for training. The internal parameters of the network are adjusted by reverse learning to improve the feature extraction accuracy. Finally, the BP neural network is used to predict the fault. Comparing the prediction results with the SVRM method, we can find that the method has certain advantages in the fault prediction for the data.