Bearing fault diagnosis method based on dimension transformation convolution depth forest

Abstract

The invention discloses a bearing fault diagnosis method based on a dimension transformation convolution depth forest, and the method comprises the following steps: carrying out the preprocessing of a bearing vibration signal, carrying out the normalization of the vibration signal, and storing a two-dimensional gray image; extracting features in the picture by using a convolutional neural network; changing the architecture of the convolutional neural network; the architecture of the deep forest is changed; replacing a SoftMax layer in the CNN with a cascaded forest structure, constructing a convolutional deep forest fault diagnosis model, inputting features extracted by the CNN with the upper layer changed, and enabling the deep forest to perform fault diagnosis by using the cascaded structure; training the cascade forest by using the training set; determining a verification index; and testing the accuracy of the model by using the test set. According to the method, the accuracy higher than that of a convolutional neural network and a deep forest can be obtained, and better accuracy can be obtained on small samples.

Description

technical field [0001] The invention belongs to the field of mechanical bearing fault diagnosis and health management (Prognostics Health Management, PHM) artificial intelligence technology, and in particular relates to a bearing fault diagnosis method based on dimension transformation convolution depth forest. Background technique [0002] As a transmission device in a mechanical system, rotating machinery is widely used in aviation machinery, agricultural machinery and modern machine tools, and plays an important role in national economic production. As an important part of the mechanical system, rotating equipment usually needs to operate under high pressure, high speed, heavy load and other environments, which greatly increases the probability of problems in rotating machinery. [0003] Bearings are one of the most critical parts in rotating machinery, and about 30% of rotating machinery failures are caused by bearing failures. The fault detection of bearing components ...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide a bearing fault diagnosis method based on dimension transformation convolution depth forest to solve the problems of difficulty in information extraction, large diagnostic errors and intermittent prediction results.

Images