Open AccessIntelligent Fault Diagnosis of Engine Based on PCA-SOM
Author(s) -
Yangyang Zhang,
Yunyi Jia,
Chiming Guo,
Weiyi Wu
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 - 10.1088/1742-6596/1453/1/012022
Subject(s) - artificial neural network , waveform , pattern recognition (psychology) , self organizing map , artificial intelligence , computer science , fault (geology) , feature extraction , dimensionality reduction , feature (linguistics) , feature vector , dimension (graph theory) , data mining , mathematics , telecommunications , radar , linguistics , philosophy , pure mathematics , seismology , geology
Self-Organizing Feature Map (SOM) is a kind of self-organizing and self-learning network without teacher. It is mainly used for pattern recognition and region classification of input vectors. A fault diagnosis method of engine fuel supply system based on SOM neural network is proposed. The sensor is used to monitor the fuel pressure waveform of a certain engine fuel supply system, time domain analysis and feature extraction are carried out on the waveform, and the feature dimension reduction is realized by PCA to form the input vector of SOM neural network. The SOM neural network is used to establish the diagnosis model and then recognize fault patterns for test samples. The results of pattern recognition show that SOM neural network can identify and classify faults accurately, and it has certain engineering application value.