Rolling bearing fault diagnosis method based on feature alignment convolutional neural network

Abstract

The invention discloses a rolling bearing fault diagnosis method based on a feature alignment convolutional neural network, and the method comprises the following steps: (1) obtaining acceleration signals of a rolling bearing in various health states, making the signals into equal-length samples, and giving a health state label; (2) dividing the samples into a training set and a test set, and setting sample weights for all the samples in the training set by taking a weight balance principle of each category; (3) building a feature alignment convolutional neural network which comprises a feature alignment structure, a feature mapping layer and a classifier, and the feature alignment structure comprises a single-step multi-scale convolution layer and a whole-period maximum pooling layer; (4)performing weighted training on the constructed network by using samples in the training set, and obtaining a trained model by using an optimization strategy; and (5) inputting the test set samples into the trained network to obtain a diagnosis result. According to the method, a feature alignment structure is adopted, so that the network model can learn more robust features, and the generalization performance of the model is improved.

Description

technical field [0001] The invention belongs to the technical field of mechanical manufacturing, and relates to a mechanical fault diagnosis technology, in particular to a rolling bearing fault diagnosis method based on a feature-aligned convolutional neural network. Background technique [0002] Rolling bearings are one of the most widely used mechanical components in modern industrial machinery and equipment. Because they usually exist as supports for various rotating machinery, they have a very important impact on the operating status of the entire mechanical equipment. The service life distribution of rolling bearings is often very discrete, and the time of failure is difficult to predict. Therefore, real-time state monitoring and fault diagnosis of bearings are of great significance to ensure the efficient operation of mechanical equipment. [0003] With the development of computer and sensor technology, the data collected in the industrial field is growing explosively....

AI Technical Summary

Problems solved by technology

However, studies have shown that small input translations can cause a cliff-like decline in the classification accuracy of convolutional neural networks, because the network is only invariant to a certain degree of translation.

In some cases, the features mentioned by the network from the translated samples are completely distorted

Images