Automatic fault diagnosis in PMSM using adaptive filtering and wavelet transform

Fault diagnosis of wind turbines is of great importance to reduce operating and maintenance costs of wind farms. At present, most wind turbine fault diagnosis methods are focused on single faults, and the methods for combined faults usually depend on inefficient manual analysis. Filling the gap, this paper proposes a low-pass filtering empirical wavelet transform (LPFEWT) machine learning based fault diagnosis method for combined fault of wind turbines, which can identify the fault type of wind turbines simply and efficiently without human experience and with low computation costs. In this method, low-pass filtering empirical wavelet transform is proposed to extract fault features from vibration signals, LPFEWT energies are selected to be the inputs of the fault diagnosis model, a grey wolf optimizer hyperparameter tuned support vector machine (SVM) is employed for fault diagnosis. The method is verified on a wind turbine test rig that can simulate shaft misalignment and broken gear tooth faulty conditions. Compared with other models, the proposed model has superiority for this classification problem.

Download Full-text

Intelligent fault diagnosis of rotating machinery based on continuous wavelet transform-local binary convolutional neural network

Knowledge-Based Systems ◽

10.1016/j.knosys.2021.106796 ◽

2021 ◽

Vol 216 ◽

pp. 106796

Author(s):

Yiwei Cheng ◽

Manxi Lin ◽

Jun Wu ◽

Haiping Zhu ◽

Xinyu Shao

Keyword(s):

Neural Network ◽

Wavelet Transform ◽

Fault Diagnosis ◽

Convolutional Neural Network ◽

Continuous Wavelet Transform ◽

Rotating Machinery ◽

Continuous Wavelet ◽

Intelligent Fault Diagnosis

Download Full-text

Fault Diagnosis of a Helical Gearbox Based on an Adaptive Empirical Wavelet Transform in Combination with a Spectral Subtraction Method

Applied Sciences ◽

10.3390/app9081696 ◽

2019 ◽

Vol 9 (8) ◽

pp. 1696 ◽

Cited By ~ 2

Author(s):

Wang ◽

Lee

Keyword(s):

Wavelet Transform ◽

Fault Diagnosis ◽

Background Noise ◽

Vibration Signal ◽

Spectral Subtraction ◽

Accurate Identification ◽

Side Band ◽

Empirical Wavelet Transform ◽

Gear Fault ◽

Fault Characteristic

Fault characteristic extraction is attracting a great deal of attention from researchers for the fault diagnosis of rotating machinery. Generally, when a gearbox is damaged, accurate identification of the side-band features can be used to detect the condition of the machinery equipment to reduce financial losses. However, the side-band feature of damaged gears that are constantly disturbed by strong jamming is embedded in the background noise. In this paper, a hybrid signal-processing method is proposed based on a spectral subtraction (SS) denoising algorithm combined with an empirical wavelet transform (EWT) to extract the side-band feature of gear faults. Firstly, SS is used to estimate the real-time noise information, which is used to enhance the fault signal of the helical gearbox from a vibration signal with strong noise disturbance. The empirical wavelet transform can extract amplitude-modulated/frequency-modulated (AM-FM) components of a signal using different filter bands that are designed in accordance with the signal properties. The fault signal is obtained by building a flexible gear for a helical gearbox with ADAMS software. The experiment shows the feasibility and availability of the multi-body dynamics model. The spectral subtraction-based adaptive empirical wavelet transform (SS-AEWT) method was applied to estimate the gear side-band feature for different tooth breakages and the strong background noise. The verification results show that the proposed method gives a clearer indication of gear fault characteristics with different tooth breakages and the different signal-noise ratio (SNR) than the conventional EMD and LMD methods. Finally, the fault characteristic frequency of a damaged gear suggests that the proposed SS-AEWT method can accurately and reliably diagnose faults of a gearbox.

Download Full-text