Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Online fault diagnosis method for rolling bearing under variable load based on transfer learning

A rolling bearing and fault diagnosis technology, which is applied in the field of online fault diagnosis of rolling bearings under variable loads of transfer learning, can solve the problems such as the inability to effectively guarantee the modeling efficiency and accuracy of online fault diagnosis of rolling bearings under variable loads, so as to solve the problems of online fault diagnosis, The effect of reducing distribution differences and high modeling efficiency

Active Publication Date: 2021-06-15
HARBIN UNIV OF SCI & TECH
View PDF14 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] In view of the above problems, the present invention proposes a transfer learning online fault diagnosis method for rolling bearings under variable loads, which is used to solve the problem that the deep migration method of the existing offline training method cannot effectively ensure the modeling efficiency and accuracy in the online fault diagnosis of rolling bearings under variable loads. rate problem

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Online fault diagnosis method for rolling bearing under variable load based on transfer learning
  • Online fault diagnosis method for rolling bearing under variable load based on transfer learning
  • Online fault diagnosis method for rolling bearing under variable load based on transfer learning

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment 1

[0087] In order to verify the validity and accuracy of the method of the present invention, the bearing data set of Case Western Reserve University in the United States is used for experimental verification [23] . Bearing test bench such as Figure 4 As shown, these data are collected by the acceleration sensor under 4 kinds of load conditions, and the sampling frequency is 12kHz.

[0088] Load types include 0hp, 1hp, 2hp and 3hp, and the motor speed varies between 1730rpm and 1797rpm depending on the load. Under each load, there are three types of fault damage levels: normal state (N) and inner ring fault (IR), outer ring fault (OR) and rolling element fault (B). The damage diameters are 0.1778mm, 0.3556mm, and 0.5334 mm respectively. mm, there is no damage to the rolling bearing under normal conditions, so there is no fault location and damage degree. Classify the normal state of rolling bearings, inner ring, outer ring, rolling element faults and 10 states of different f...

specific Embodiment 2

[0117] In order to better verify the efficiency of the method of the present invention, the time consumption and accuracy of the three methods were compared. The setting of the online data set is consistent with Table 5, and the experimental results are shown in Table 7.

[0118] Table 7 Comparison of time consumption and accuracy of the three methods

[0119]

[0120] In the experiment, method 1 and method 2 adopt the offline training method, that is, retrain the model each time to process the new online data; method 3 is the online training method, and only need to update the model for the newly added online data. Method 1 is to combine CNN and Softmax to carry out backpropagation optimization training CNN model; method 2 is to carry out feature extraction to the fault data with the CNN model trained in method 1, and then input SVM to carry out fault classification; method 3 is the CNN proposed by the present invention -ISVM rolling bearing multi-state identification meth...

specific Embodiment 3

[0123] In order to further prove that the method of the present invention has obvious advantages in online fault diagnosis of rolling bearings under variable loads, ISVM, OS-ELM [28] And the online learning method proposed in the literature [29] for comparative experimental research. Among them, the ISVM and OS-ELM methods adopt the feature extraction method of literature [30] to extract 10 kinds of time-domain features and 4 kinds of frequency-domain features, and then input the extracted features into the diagnosis algorithm to complete the fault diagnosis. During the experiment, the same source domain and target domain datasets were used, the time comparison results are shown in Table 8, and the accuracy comparison results are shown in Table 8. Figure 11 shown.

[0124] Table 8 Time comparison with other methods

[0125]

[0126] It can be seen from Table 8 that the average online learning time of the method proposed in the present invention is 16.97s, which is about ...

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

No PUM Login to View More

Abstract

The invention discloses an online fault diagnosis method for a rolling bearing under a variable load based on transfer learning, which belongs to the technical field of fault diagnosis and is used for solving the problem that the modeling efficiency and accuracy in online fault diagnosis of a rolling bearing under a variable load cannot be effectively ensured by a depth transfer method of an existing offline training mode. The method is technically characterized by comprising the following steps: firstly, performing STFT processing on an original time domain vibration signal, and constructing a two-dimensional spectrum data set; then training a source domain CNN-ISVM model by using source domain data to obtain a source domain classification model, storing model parameters and migrating the model parameters to a target domain CNN-ISVM training process; and finally, updating and correcting an ISVM classifier in the target domain CNN-ISVM model through online data to realize multi-state online recognition of a rolling bearing under a variable load. According to the method, the model training time is greatly shortened, the calculation amount is greatly reduced, the modeling efficiency is high, and meanwhile the accuracy rate and the generalization performance are high. The method has important guiding significance for online monitoring and rapid diagnosis of faults of a rolling bearing in actual work.

Description

technical field [0001] The invention relates to the technical field of fault diagnosis, in particular to an online fault diagnosis method for rolling bearings under variable load through transfer learning. Background technique [0002] As a key component in industrial production equipment, rolling bearings will directly affect the safe operation of the entire mechanical equipment once they fail [1] . Under variable load conditions, rolling bearings have complex and changeable operating states, and the distribution of data features is quite different. In practice, the target domain data collection is often obtained online in a sequential manner, which cannot satisfy the completeness assumption of the test data in the model training process, and poses new challenges for the existing transfer learning methods [2] . Therefore, it is of great research significance to carry out multi-state online fault diagnosis of rolling bearings under variable loads. [0003] With the deepe...

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
no application Login to View More
IPC IPC(8): G01M13/045G06N3/04G06N3/08G06K9/62
CPCG01M13/045G06N3/084G06N3/047G06N3/045G06F18/2411G06F18/2415
Inventor 康守强刘旺辉王玉静王庆岩梁欣涛谢金宝
Owner HARBIN UNIV OF SCI & TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products