Fault diagnosis of wind turbine gearbox based on wavelet neural network

Abstract

In order to monitor the wind turbine gearbox running state effectively, a fault diagnosis method of wind turbine gearbox is put forward based on wavelet neural network. Taking a 1.5 MW wind turbine gearbox as the target of study, the frequency spectrum of vibration signal and the fault mechanism of driving part are analyzed, and the eigenvalues of the frequency domain are extracted. A wavelet neural network model for fault diagnosis of wind turbine gearbox is established, and wavelet neural network is trained by using different feature vectors of fault types. The relationship between fault component and vibration signal is identified, and the vibration fault of wind turbine gearbox is predicted and diagnosed by network model. The analysis results show that the method can diagnose fault and fault pattern recognition of wind turbine gearbox very well.

Machine learning for fault analysis in rotating machinery: A comprehensive review

“…6 , most studies developed an intelligent model to find the fault types of a single component. However, there are a considerable amount of studies that also considered multi-component faults to build IFDP models for rotating machines [ [91] , [92] , [93] , [94] , [95] , [96] , [97] , [98] , [99] , [100] , [101] ]. Fig.…”

Investigation of Isolation Forest for Wind Turbine Pitch System Condition Monitoring Using SCADA Data

“…The trend of time-domain features was examined and the root mean square (RMS) of the signal was found to trend upwards before failure for all cases. Neural networks have also been used in anomaly detection, Huitao et al [31] utilised a wavelet neural network for a test rig gearbox. The network outputs a range of probabilities and the input data corresponds to a specific fault out of five possible faults.…”

Comparison of New Anomaly Detection Technique for Wind Turbine Condition Monitoring Using Gearbox SCADA Data

“…This test rig had faults introducted artificially to it, and one second measurements were taken six times over 7 h. The trend of the time domain features was examined and it was found the root mean square (RMS) of the signal showed a trend upwards towards failure for all cases considered. Huitao et al [10] presented a technique utilising a wavelet neural network for a test rig gearbox. The network outputs a range of probabilities that the input data corresponds to a specific fault out of five possibilities, and was shown to perform better than Empirical Model Decomposition (EMD).…”