Rolling bearing residual life prediction method based on convolutional long-short-term memory recurrent neural network

Abstract

The invention provides a rolling bearing residual life prediction method based on a convolution long-short-term memory recurrent neural network, and the method comprises the steps: obtaining a vibration acceleration signal of a rolling bearing, and enabling the vibration acceleration signal to serve as a data sample; establishing a convolution auto-encoder, taking the data sample as an input, andtraining to obtain a multi-layer convolution auto-encoder; obtaining feature data of the data sample according to the data sample and the multi-layer convolution auto-encoder; determining a health evaluation index of the rolling bearing by using a fuzzy C-means algorithm according to the feature data of the data sample; establishing a convolutional long-short-term memory recurrent neural network model, taking the recession period bearing data sample and the corresponding health evaluation index as input of the convolutional long-short-term memory recurrent neural network, and training to obtain a prediction model; according to the method, the dependence on artificial experience and professional knowledge is reduced, the data volume of the input long-short-term memory recurrent neural network is effectively reduced, the operation is simplified, the prediction model is comprehensive, and the recession state of the bearing is effectively reflected.

Description

technical field [0001] The invention relates to the technical field of rolling bearing life prediction, in particular to a method for predicting the remaining life of rolling bearings based on a convolutional long-short-term memory cycle neural network. Background technique [0002] As an important part of modern manufacturing, bearings play a vital role in the reliable operation of most rotating machinery, directly affecting the health of the mechanical system. During the operation of mechanical equipment, changes and uncertainties in complex working conditions will cause damage to rolling bearings, directly affecting the accuracy and reliability of the entire mechanical equipment system. By effectively monitoring the working state of rolling bearings and accurately predicting their degradation trends, it is possible to ensure the safe and reliable operation of mechanical equipment and avoid major economic losses. Therefore, it is necessary to monitor the health of the bea...

AI Technical Summary

Problems solved by technology

However, these feature quantities are only sensitive to a certain stage of the bearing degradation process. If a certain feature quantity is used alone as the degradation feature quantity of the bearing, the error of the prediction result is large

The comprehensive index of multiple features can describe the bearing degradation process relatively accurately, but increases the computational complexity

In addition, in most of the existing prediction models, the extraction of features still requires expert experience and manual extraction.

Images