Fault feature screening method based on weighted multi-feature fusion and SVM classification

Abstract

The invention provides a fault feature screening method based on weighted multi-feature fusion and SVM classification. The method comprises the following steps of 1, obtaining the time series data; 2,extracting a time domain (T), a frequency domain (P), the energy (E) and an entropy feature (S), and forming a high-dimensional feature set (Q); 3, screening out the features (Q1) with the diagnosisrate greater than 50%, carrying out the correlation analysis, and removing the features (Q2) with the similarity greater than 85%; 4, selecting a feature with the highest score through PCA and a loadscoring method to form a new sub-feature set (T3, P3, E3, S3); 5, carrying out SVM diagnosis on the T3, the P3, the E3 and the S3, and obtaining a weight Wi according to the diagnosis rate Ri; 6, performing the weighted fusion of the features; and 7, inputting the fused features into a classifier for diagnosis. Through the above steps, a group of optimal features capable of maintaining the fault intrinsic information is obtained, the original failure information represented by the features is ensured, the fault diagnosis accuracy is improved, and the method is of great significance to the efficient mechanical fault diagnosis.

Description

technical field [0001] The invention relates to a fault feature screening method based on weighted multi-feature fusion and SVM classification, which uses the time domain, frequency domain, energy and entropy features of vibration signals to form a specific fault feature set, combined with support vector machine (SVM) and main Methods such as component analysis (PCA) screen features to find key features that can reflect fault information and failure forms, and perform weighted fusion of different dimensional features according to the screening results. It is suitable for technical fields such as signal processing and mechanical fault diagnosis. Background technique [0002] Rotating machinery is a very important power device for modern industrial applications, and rolling bearings are important components for power transmission in rotating machinery, and their operating status is directly related to the performance status of mechanical equipment. Once a failure occurs, it m...

AI Technical Summary

Problems solved by technology

A single aspect of the feature cannot fully reflect the internal information of the bearing fault, and the increase in the dimension of the feature, although the information of the bearing is more comprehensive, but the increase in the dimension increases the invalid information on the one hand, resulting in an increase in the complexity of the diagnosis process, and on the other hand Different types of faults have different applicability in different stages of bearing operation, which ultimately reduces the accuracy of diagnostic results

[0004] Therefore, after the feature extraction is finished, the features should be evaluated, and the best features that can maintain the intrinsic information about the fault should be retained while reducing the number of them as much as possible, so that the fault diagnosis of rotating machinery can be carried out effectively and efficiently. At present, the research on this aspect is still relatively weak, and most of them only use a single method, such as directly using PCA to screen features, which cannot reflect the differences between samples, and most of these methods reduce multi-dimensional features to low-dimensional by means of compression. The original failure information represented by the feature is lost

Images