CN-109858104-B - Rolling bearing health assessment and fault diagnosis method and monitoring system

CN109858104BCN 109858104 BCN109858104 BCN 109858104BCN-109858104-B

Abstract

The invention discloses a rolling bearing health assessment and fault diagnosis method and a rolling bearing health assessment and fault diagnosis monitoring system, which solve the problem that a large amount of prior known data or too much manual experience intervention is needed to ensure the monitoring effect in the prior art, and have the effect of accurately detecting and identifying the bearing fault by carrying out online real-time analysis on a bearing vibration signal; the technical scheme is as follows: the method comprises the following steps: obtaining a vibration signal of a bearing, and processing the vibration signal to obtain a spectrogram; establishing a graph model for the spectrogram; similarity comparison is carried out on the adjacent matrixes generated by the graph model to calculate the degree of abnormality, and decision is carried out on the degree of abnormality indexes; setting a threshold value for hypothesis testing, and carrying out fault testing on the bearing; and carrying out fault diagnosis when the bearing signal is in fault.

Inventors

LU GUOLIANG
ZHANG DI
YAN PENG

Assignees

山东大学
山东大学

Dates

Publication Date: 20220902
Application Date: 20190110
Priority Date: 20190110

Claims (8)

1. A rolling bearing health assessment and fault diagnosis method is characterized by comprising the following steps: step (1) obtaining a vibration signal of a rolling bearing, and processing the vibration signal to obtain a spectrogram; step (2) establishing a graph model for the spectrogram; in the step (2), selecting a frequency interval, dividing the frequency interval into frequency segments with equal length, and calculating the energy of each frequency segment; each frequency bin is taken as a graph structure vertex,the connecting line between two frequency bands is used as the weighting edge of the graph structure, and the difference value of the energy of each frequency band is used as the weighting weight d of the weighting edge i,j Wherein i and j are any two points in the vertex; step (3) similarity comparison is carried out on the adjacency matrixes generated by the graph model to calculate the degree of abnormality, and decision is carried out on the degree of abnormality indexes; setting a threshold value for hypothesis testing, and carrying out fault testing on the rolling bearing; and carrying out fault diagnosis when the bearing signal is in fault.
2. The rolling bearing health assessment and fault diagnosis method according to claim 1, wherein in the step (1), a window function is selected, and the collected vibration signal is subjected to windowing processing; and carrying out Fourier transform on the vibration signal in the window to obtain a spectrogram.
3. The rolling bearing health assessment and fault diagnosis method according to claim 1, characterized in that the weight d is set i,j And the ith row and the jth column in the matrix are used as numerical values, so that the graph structure is converted into an adjacent matrix of N x N, wherein N is the number of the frequency segments.
4. Rolling bearing health assessment and fault diagnosis method according to claim 1, characterized in that in said step (3), an adjacency matrix X is paired t Performing diagonalization decomposition to calculate an abnormality degree s t And the degree of abnormality of the adjacent matrix is decided through martinggle-test.
5. The rolling bearing health assessment and fault diagnosis method according to claim 1, wherein in the step (4), if the bearing signal is normal, the average value of the graph model at the current time and the graph model at the previous time is used as a new graph model, and fault detection of data at the next time is performed.
6. The rolling bearing health assessment and fault diagnosis method according to claim 5, wherein if a bearing signal fails, an alarm is given and fault diagnosis is performed; and selecting fault signals of different fault types, calculating the weight of each row of the adjacent matrix of the graph model by an entropy method, and inputting the weight as a feature vector into the SVM for training.
7. The rolling bearing health assessment and fault diagnosis method according to claim 6, wherein the weight of each row of the adjacency matrix of the fault time graph model is calculated and input into the SVM for fault diagnosis.
8. A monitoring system for health assessment and fault diagnosis of a rolling bearing is characterized by comprising an acceleration sensor, a computer readable storage medium and a processor, the acceleration sensor is used for monitoring a vibration signal of the rolling bearing in the running process and transmitting the vibration signal to the processor; a computer-readable storage medium stores a computer program which, when executed by a processor, implements the method of any one of claims 1-7.

Description

Rolling bearing health assessment and fault diagnosis method and monitoring system Technical Field The invention relates to the field of online monitoring of faults of rolling bearings, in particular to a method and a system for health assessment and fault diagnosis of a rolling bearing. Background The rolling bearing is used as a basic part of a rotary machine, and the working state of the rolling bearing has a great influence on the safety of the whole equipment and the whole production line. Therefore, it is of great significance to carry out fault diagnosis on the system. However, the rolling bearing signal has the characteristics of nonlinearity and non-stationarity, and the fault characteristics are difficult to find only from the time domain and the frequency domain. The appearance of time-frequency methods (such as short-time fourier transform, wavelet packet decomposition, etc.) effectively remedies this deficiency. Although the existing method also achieves certain effect, a large amount of prior known data or excessive human experience intervention is generally required to ensure the monitoring effect. Disclosure of Invention In order to overcome the defects of the prior art, the invention provides a rolling bearing health assessment and fault diagnosis method and a rolling bearing health assessment and fault diagnosis monitoring system, which have the effects of accurately detecting and identifying the faults of the rolling bearing by carrying out online real-time analysis on the vibration signals of the rolling bearing. The invention adopts the following technical scheme: the rolling bearing health assessment and fault diagnosis method comprises the following steps: step (1) obtaining a vibration signal of a rolling bearing, and processing the vibration signal to obtain a spectrogram; step (2) establishing a graph model for the spectrogram; step (3) similarity comparison is carried out on the adjacent matrixes generated by the graph model to calculate the degree of abnormality, and decision is carried out on the degree of abnormality indexes; setting a threshold value for hypothesis testing, and carrying out fault testing on the rolling bearing; and carrying out fault diagnosis when the bearing signal is in fault. Further, in the step (1), a window function is selected, and windowing processing is performed on the acquired vibration signal; and carrying out Fourier transform on the vibration signal in the window to obtain a spectrogram. Further, in the step (2), a frequency interval is selected and divided into frequency segments with equal length, and the energy of each frequency segment is calculated. Furthermore, each frequency segment is taken as a vertex of the graph structure, a connecting line between two frequency segments is taken as a weighted edge of the graph structure, and the difference of the energy of each frequency segment is taken as the weight d of the weighted edge i,j Wherein i and j are any two points in the vertex. Further, the weight d is set i,j And the number of the ith row and the jth column in the matrix is used for converting the graph structure into an adjacent matrix of N by N, wherein N is the number of the frequency segments. Further, in the step (3), the adjacency matrix X is subjected to t Performing diagonalization decomposition to calculate an abnormality degree s t And the degree of abnormality of the adjacent matrix is decided through martinggle-test. Further, in the step (4), if the bearing signal is normal, the average value of the graph model at the current time and the graph model at the previous time is used as a new graph model, and fault detection of data at the next time is performed. Further, if the bearing signal fails, alarming is carried out, and fault diagnosis is carried out; and selecting fault signals of different fault types, calculating the weight of each row of the adjacent matrix of the graph model by an entropy method, and inputting the weight as a feature vector into the SVM for training. Further, the weight of each row of the adjacent matrix of the fault moment graph model is calculated and input into the SVM for fault diagnosis. A monitoring system for health assessment and fault diagnosis of a rolling bearing comprises an acceleration sensor, a computer readable storage medium and a processor, the acceleration sensor is used for monitoring a vibration signal in the running process of the bearing and transmitting the vibration signal to the processor; the computer-readable storage medium stores a computer program that, when executed by a processor, implements the bearing health assessment and fault diagnosis method described above. Compared with the prior art, the invention has the beneficial effects that: (1) according to the invention, fault detection is carried out on the vibration signal of the rolling bearing during the operation of the machine, and then the health condition of the rolling bearing is evaluated, so that the real