In-depth adversarial diagnosis method for fan bearing faults in unbalanced small sample scenarios
A small-sample, unbalanced technique, applied in mechanical bearing testing, neural learning methods, computer components, etc.
Active Publication Date: 2020-12-18
NORTHEAST DIANLI UNIVERSITY
View PDF9 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004] The purpose of the present invention is to overcome the deficiencies of the prior art, to provide a scientific and reasonable, strong adaptability, high practical value, capable of high noise interference, insufficient number of samples and unbalanced training set scale of different types of samples, and complex scenarios, In-depth adversarial diagnosis method for fan bearing faults in unbalanced small sample scenarios with better fault identification accuracy
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0054] In the present invention, a fan bearing fault depth confrontation diagnosis method in consideration of unbalanced small sample scenarios includes the following steps:
[0055] 1) Wind turbine bearing vibration signal collection
[0056] Acceleration sensors are used to collect wind turbine bearing normal state signals, rolling element fault vibration signals, inner ring fault vibration signals and outer ring fault signals, and record the above signals with a 16-channel data recorder, and the signal sampling frequency is 12kHz;
[0057] 2) Improve AC-GAN model construction
[0058] Generative Adversarial Networks (GAN) consists of two parts: Generator (G) and Discriminator (D); G maps the noise signal z to the sample space to obtain the generated sample data X fake =G(z); will generate sample X fake or real sample X real Input the discriminator, judge by D and output the probability value (P(S|X)=D(X)), which indicates the probability of discriminating the sample X be...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The present invention is a deep confrontation diagnosis method for wind turbine bearing faults in an unbalanced small sample scenario, which is characterized in that it includes wind turbine bearing vibration signal collection, improved AC-GAN model construction, improved AC-GAN sample construction, and wind turbine bearing vibration signals Sample generation and wind turbine bearing fault diagnosis steps in various scenarios: solve the problems of complex vibration signal and noise interference, few fault samples and unbalanced number of samples between categories in the fault diagnosis of wind turbines based on vibration signals, and improve the performance of small sample unbalanced scenarios. The accuracy of fault identification is low, and it can have a good fault identification accuracy in complex scenarios such as high noise interference, insufficient number of samples, and unbalanced training set scale of different types of samples. It is scientific and reasonable, strong adaptability, and high practical value. Provide reference for wind turbine research and development, wind farm operation and maintenance, wind turbine research and other related personnel.
Description
[0001] The invention relates to a fault diagnosis method for fan bearings, in particular, a deep confrontation diagnosis method for fan bearing faults in an unbalanced small-sample scenario, which is applied to online diagnosis of mechanical fault states of wind turbine bearings. Background technique [0002] Bearings, as the core components of the wind turbine transmission system, are continuously affected by alternating impact forces and loads, and become parts with high incidence of mechanical failures. Once the bearing is damaged, the downtime of the fan will be long, the repair cost will be high, and the economic loss will be serious. Therefore, accurate diagnosis of wind turbine bearing faults is of great significance to ensure the safe and reliable operation of wind turbines and the economy of wind farms. Existing fan bearing fault diagnosis is generally carried out based on bearing vibration signals, and the diagnosis process is divided into two parts: feature extracti...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G01M13/045G01M13/04G06K9/00G06K9/62G06N3/04G06N3/08
CPCG01M13/045G01M13/04G06N3/082G06N3/045G06F2218/08G06F2218/12G06F18/24133Y02B10/30
Inventor 黄南天杨学航蔡国伟宋星王文婷陈庆珠赵文广刘宇航刘德宝包佳瑞琦张祎祺吴银银
Owner NORTHEAST DIANLI UNIVERSITY