A Fault Diagnosis Method for Rolling Bearings Based on Bidirectional Memory Recurrent Neural Network

Abstract

The invention discloses a rolling bearing fault diagnosis method based on a two-way memory cyclic neural network, aiming at that the existing rolling bearing fault diagnosis method does not consider the single logical structure of the data after feature extraction and cannot judge the fault type from the data as a whole when processing the fault data Firstly, obtain program data samples, standardize and preprocess the vibration acceleration data, so that the collected data conform to the standard normal distribution, and then use the time-frequency domain feature extraction algorithm to obtain 512 time-frequency domain feature vectors; then, construct an improved The two-way memory-type cyclic neural network fault diagnosis model adopts the idea of ​​simple design. The designed memory-type cyclic neural network includes forgetting gate, input gate and cell state, and then uses sample data to train the neural network weight parameters. After training Iterate to generate a model that maps the relationship between bearing data and failure types; finally use this model for failure analysis, enabling accurate diagnosis of rolling bearing failures.

Description

technical field [0001] The invention relates to a fault diagnosis method of a rolling bearing, in particular to a fault diagnosis method of a rolling bearing based on a bidirectional memory cyclic neural network. Background technique [0002] With the rapid development of modern industry and science and technology, production equipment is gradually developing towards high speed, automation and intelligence, and problems such as aging of equipment components and changes in application environment are often unavoidable. The aging of components will lead to mechanical equipment. Various faults occur during operation, resulting in huge economic losses. Therefore, it is of great significance to collect massive data for equipment status monitoring, and to research and use innovative theories and methods to efficiently mine information from equipment big data for equipment status fault diagnosis. . [0003] Rolling bearings are the core components of rotating machinery. The fault ...

AI Technical Summary

Problems solved by technology

[0004] The design of the traditional fault diagnosis model does not take into account the unsteady characteristics of the bearing during data collection. It is necessary to design a bearing fault diagnosis method for this feature to improve the accuracy of rolling bearing fault diagnosis. At the same time, due to structural limitations, the length of single processing data is limited in traditional methods (one sample data needs to be divided into multiple batches for processing ), cannot fully combine the data as a whole to judge the fault type

Images