Multiple super-order analysis method for vibration signal feature extraction

Abstract

The invention discloses a multiple super-order analysis method for vibration signal feature extraction. The multiple super-order analysis method for vibration signal feature extraction comprises the following steps that 1, an extreme value of a to-be-processed original time series is acquired; 2, an extreme value increment series of the original series is calculated; 3, a mean value removing series of the time series is constructed; 4, the mean value removing series is divided into 2Ns intervals in the positive and negative directions according to the time scale length; 5, high-order polynomial tendency fitting is conducted on each interval; 6, a mean square error of data of each interval is calculated; 7, fluctuation functions are calculated for all the intervals; 8, the time scale is changed, the steps from 4 to 7 are repeated, and a generalized Hearst index is acquired; and 9, a singular index and a multiple fractal singular spectrum are calculated, and feature points are extractedto serve as signal features. The multiple super-order analysis method for vibration signal feature extraction has the advantages that fluctuation and impact features in original signals are highlighted, and the multiple fractal features of the nonstationary time series are effectively analyzed.

Description

technical field [0001] The invention relates to the technical field of processing vibration signals of rotating machinery, in particular to a multiple super-order analysis method for feature extraction of vibration signals. Background technique [0002] The fault diagnosis method mainly focuses on the operating state of the system, which can detect faults in time and guide maintenance, which plays an important role in improving system reliability. Generally, when a rotating component fails, the damage point will produce an impact in the load area and contact with other components, which will aggravate the vibration of the component. Therefore, the vibration signal carries important diagnostic information and is an important basis for equipment status identification. Simultaneous vibration usually shows non-stationarity and multifractal characteristics in the signal. Therefore, extracting characteristic information that can characterize faults from complex fault signals has ...

AI Technical Summary

Problems solved by technology

However, for similar types of faults, the MF-DFA method has the problems of close extraction parameters, crossover and different degrees of state aliasing, which affect the accuracy of fault feature extraction.

Images