Spindle bearing fault detection method and system, equipment and readable storage medium

Abstract

The invention discloses a spindle bearing fault detection method and system, equipment and a readable storage medium. The method comprises the following steps that: based on a deep learning wide residual network and a signal processing technology, ensemble empirical mode decomposition is performed on vibration signals, mode components are screened based on a kurtosis value, singular value decomposition optimization is performed on screened-out components; signals are reconstructed; signal features in an image form are outputted through short-time Fourier transform; a wide residual network us built;the image feature data are inputted into the wide residual network to train the network;and finally the wide residual network with a fault diagnosis function is obtained. According to the method, signal processing technology and deep leaning technology are combined, so that the influence of bearing vibration signal noise is reduced, meanwhile; the wide residual network also relieves the requirement of a general artificial neural network model for the high performance of a computer; the problem of performance degradation frequently occurring along with the increase of the number of network layers is solved; and the accuracy and efficiency of fault diagnosis are improved.

Description

technical field [0001] The invention belongs to the field of spindle bearing fault diagnosis, and in particular relates to a spindle bearing fault detection method, system, equipment and a readable storage medium. Background technique [0002] With the continuous development of modern machinery and equipment, large-scale, precision, and complexity have become the trend. The service conditions of equipment are varied, and are often accompanied by harsh working conditions such as variable load, noise interference, and impact. Bearings are the key components of the main shaft. According to relevant data, about 30%-40% of equipment failures occur on bearings. If these failures cannot be found in time, it will cause serious economic losses and even cause safety accidents. Therefore, it is very important to monitor the state of the main shaft bearing and carry out timely fault diagnosis and identification, which not only guarantees the safe service of mechanical equipment, but als...

AI Technical Summary

Problems solved by technology

[0005] The traditional convolutional neural network obtains stronger analysis ability by increasing the number of layers, but as the depth increases, the phenomenon of gradient explosion or gradient disappearance becomes more and more serious, and the network degradation becomes more obvious, which seriously affects the training effect of the network. , but the effect of shallow network on the extraction and identification of fault features is poor, so it is necessary to set the number of layers of the network reasonably

At the same time, because the vibration signal of the bearing usually contains strong noise, the noise is random and complex, and the deep learning model usually "sees" the noise as the information to be learned together for training, which will result in extremely low training efficiency and Accuracy

Images