A fault signal analysis method based on complex wavelet modulus maxima connection line

Abstract

The invention discloses a fault signal analysis method based on a complex wavelet modulus maximum connecting line, which comprises steps that: the fault signal is obtained and processed by continuouswavelet transform; the vanishing moment of complex wavelet is optimized and the vanishing moment of complex wavelet is judged to be optimal; then the order n of vanishing moment is chosen to be greater than Lipschitz exponent alpha and the complex wavelet transform is carried out; the modulus maxima of continuous wavelet transform coefficients are plotted; the singular point x0 is determined alongthe modulus maxima connection line, and the corresponding slope of log2 | Wf (s, k) | along the modulus maxima connection line converging to x0 is calculated, which slope is alpha + 1 / 2, so that theLipschitz exponent alpha at x0 point is obtained. Compared with the prior art, the invention has strong practical function, simple steps and can be extended to the singularity analysis of other signals.

Description

technical field [0001] The invention relates to the technical field of fault signal analysis, in particular to a fault signal analysis method based on complex wavelet modulus maxima connection. Background technique [0002] As an effective and powerful mathematical analysis tool, wavelet transform has a good application prospect in the analysis of transient signals. At present, the real wavelet is the first choice for ranging. Of course, the real wavelet still has many advantages, such as compact support, orthogonality, high vanishing moment, accurate reconstruction, approximate symmetry, etc., but its defects mainly have three major aspects: translation sensitivity, direction Poor performance and lack of phase information. [0003] When analyzing the signal, translation sensitivity, poor directionality, and unresolved frequency domain will cause distortion of the transformation result, and the singularity of the signal cannot be accurately extracted due to oscillation. Th...

AI Technical Summary

Problems solved by technology

At present, the real wavelet is the first choice for ranging. Of course, the real wavelet still has many advantages, such as compact support, orthogonality, high vanishing moment, accurate reconstruction, approximate symmetry, etc., but its defects mainly have three major aspects: translation sensitivity, direction Poor sex, lack of phase information

[0003] When analyzing signals, translation sensitivity, poor directionality, and unresolved frequency domain will cause distortion of the transformation results, and the singularity of the signal cannot be accurately extracted due to oscillations

The translation sensitivity of real wavelet will lead to poor effect of real wavelet analysis on translation signal

Due to the uncertainty of intercepting the actual signal, it is difficult to grasp the time of intercepting the signal, especially when selecting the power signal, there is no reasonable theoretical basis for how to select the signal point, so that the signal selection is random

If the real wavelet is used to analyze the signal, the accuracy of the result is unpredictable, especially for the signal that needs to be analyzed multiple times. Since the selection of signal points is different when analyzing the fault signal, there is a certain difference in each analysis result. The result It is difficult to judge the difference

Images