A Gear Fault Diagnosis Method Based on Scalable Displacement Hyperdisk

Abstract

The invention discloses a gear fault diagnosis method based on a telescopic displacement ultra-disc. The specific steps are: measuring vibration signals of objects in different working states or fault types; extracting features commonly used in gear fault diagnosis from the vibration signals ; Divide the eigenvalues ​​of different working states into training samples and test samples; use the training samples to train the scalable shift hyperdisc classifier to establish the optimal diagnostic model; use the optimal diagnostic model to classify the test samples; according to The classification results identify the working state or fault type of the object. The gear fault diagnosis method based on the telescopic shifting ultra-disk proposed by the invention has a high recognition degree in the pattern recognition process.

Description

technical field [0001] The invention belongs to the field of gear fault diagnosis, in particular to a gear fault diagnosis method based on a telescopic shifting ultra-disc. Background technique [0002] Reasonable and effective fault diagnosis of rotating machinery is the key to ensure its reliability. As a key component, gears have been widely used in rotating machinery, so it is of great practical significance to study gear fault diagnosis methods. Among them, the intelligent diagnosis method, as an important research content in the field of fault diagnosis, has been widely concerned by researchers. [0003] Intelligent fault diagnosis is essentially a pattern recognition problem. Commonly used pattern recognition methods include methods based on probability and statistics, neural networks and support vector machines (SVM). When applied to gear fault diagnosis, the method based on probability and statistics needs enough fault training samples to make the estimated probab...

AI Technical Summary

Problems solved by technology

However, the existence of outliers will affect the classification accuracy due to the simpler geometric model of hyperdisk

And the degree of compactness or looseness that cannot be adjusted affects a more reasonable estimate of the class distribution

Images