Method and device for detecting motor bearing failure

Abstract

The invention provides a method and a device for detecting motor bearing failure. The method comprises the following steps of: collecting a vibration signal of the motor bearing; decomposing the collected vibration signal into a plurality of modal signal components by using a variational mode decomposition method; determining a sample entropy value of the vibration signal based on the decomposed plurality of modal signal components; and determining the sample entropy value of the vibration signal as a fault characteristic of the motor bearing to determine a fault type of the motor bearing. Thedevice comprises a signal collection unit, a signal decomposition unit, an entropy value determination unit, and a failure determination unit, wherein the signal collection unit collects the vibration signal of the motor bearing; the signal decomposition unit decomposes the collected vibration signal into a plurality of modal signal components by using a variational mode decomposition method; theentropy value determination unit determines the sample entropy value of the vibration signal based on the decomposed plurality of modal signal components; and the fault determination unit determinesthe sample entropy value of the vibration signal as the fault characteristic of the motor bearing to determine the fault type of the motor bearing.

Description

technical field [0001] The present application relates to the field of fault diagnosis, in particular to a method and device for detecting motor bearing faults based on variational mode decomposition (VMD), sample entropy and multi-classification correlation vector machines. Background technique [0002] In recent years, with the large-scale and complex development of modern machinery and equipment, the normal operation of equipment is of great significance to ensure the safety and efficiency of the manufacturing process. As the main drive equipment, the motor has become one of the indispensable parts of mechanical equipment or rotating system. Rolling bearings are the most important part of motors, and bearing failures account for about 50% of motor failures. Once a bearing failure occurs, it may lead to the shutdown of the entire production line, and even cause personal injury or death of the staff in severe cases. Therefore, it is of great significance to study the faul...

AI Technical Summary

Problems solved by technology

When the motor bearing fault occurs, the complex working environment and working mechanism of the mechanical transmission motor make the harmonic signal of the bearing fault generally manifest as non-stationary and nonlinear, which makes it difficult to accurately extract the fault characteristics and the signal-to-noise ratio is low, which in turn affects the motor. Status recognition rate of bearing faults

Existing various fault diagnosis techniques (such as Empirical Mode Decomposition (EMD)) use the method of envelope calculation, which makes the signal components obtained by decomposition discontinuous. After multiple decompositions, the envelope estimation error is reduced by There is serious mode aliasing phenomenon in the amplified and decomposed signal, which leads to insufficient feature extraction

In addition, after feature extraction in different fault states, existing fault diagnosis techniques usually use binary classifiers such as support vector machines (SVM) and correlation vector machines (RVM) to identify fault states , so the existing various fault diagnosis technologies are relatively simple in identifying fault types, and it is difficult to accurately distinguish and identify minor faults and inherent defects of the motor

Images