Intelligent Fault Diagnosis Method Based on Attention Mechanism Meta-learning Model in Small Samples

Abstract

The invention discloses a small-sample fault intelligent diagnosis method based on the attention mechanism meta-learning model. The invention uses the attention mechanism and the meta-learning method to establish an associated network model, performs short-time Fourier transform on mechanical signals, and obtains its time spectrum diagram , and further feature extraction and operating status identification can effectively mine the rich fault information contained in it. The method of the invention can self-adaptively train a pseudo-distance to evaluate the similarity between related data, does not need a clear mathematical formula definition, and can obtain a higher accuracy rate of mechanical equipment fault diagnosis. In this way, the dependence of the feature extraction process in the traditional diagnosis method on manual experience and the dependence of the existing intelligent fault diagnosis algorithm on a large amount of training data can be overcome, and the problem of mechanical equipment fault diagnosis under the condition of small sample data can be effectively solved.

Description

technical field [0001] The invention relates to the field of fault diagnosis of mechanical equipment, in particular to an intelligent fault diagnosis method based on a meta-learning model of an attention mechanism under small samples. Background technique [0002] Due to the rapid development of computer science and communication technology, a large amount of industrial field data has been recorded and preserved. However, the amount of data containing rich fault information is very small, and most of them are useless data. On the other hand, the artificial fault data collected in the laboratory is difficult to completely simulate the occurrence of real faults. At the same time, it is even more difficult to obtain data on the natural development of mechanical equipment faults, which requires a lot of manpower and material resources. The above problems contradict the preconditions of the existing intelligent fault diagnosis algorithms, because the existing intelligent fault d...

AI Technical Summary

Problems solved by technology

It can be seen that the small sample problem has seriously affected the timeliness, effectiveness and versatility of intelligent fault diagnosis algorithms for fault diagnosis and condition monitoring of mechanical equipment. Research on Fault Diagnosis Method of Mechanical Equipment under Low Level Problems

Images