Ultrasonic detection defect qualitative identification method based on neural network

Abstract

The invention discloses an ultrasonic detection defect qualitative recognition method based on a neural network, and the method comprises the steps: carrying out the preprocessing of a damage signal through employing a wavelet packet threshold noise reduction algorithm in a wavelet analysis algorithm, reserving a useful signal in a first intrinsic mode component as much as possible, and carrying out the mode decomposition of the signal through employing a complementary set empirical mode decomposition algorithm; carrying out soft threshold noise reduction and rigrsure noise reduction, finally, carrying out superposition reconstruction on two parts of processed intrinsic mode components to obtain a final signal, and secondly, extracting feature vectors of different damage conditions to form a learning sample of a multivariable interpolation radial basis function. According to the method, noise reduction processing can be carried out on the collected signals, the convergence speed is high, the method is simple and effective, the radial basis function neural network after learning training has the capacity of ultrasonic detection defect qualitative recognition, device damage and the damage degree can be accurately recognized, and the damage positioning can be achieved.

Description

technical field [0001] The invention relates to a qualitative recognition method for ultrasonic detection defects based on a neural network, in particular to the processing of echo signals and the damage detection of devices, and belongs to the technical field of damage signal recognition and processing. Background technique [0002] With the development of modern industry, many equipment, devices and other products have become more sophisticated, and their production and processing processes have become more complicated, and their technical parameters are often not precisely controlled, which will cause certain defects inside and on the surface of the product, which will affect the quality of the product. Product performance and even safety. Therefore, the key to the safe application of the product lies in the reasonable detection of internal and surface defects, and avoiding the use of products with potentially dangerous defects. Usually, people will conduct non-destructiv...

AI Technical Summary

Problems solved by technology

Conventional Fourier analysis theory has limitations in time-frequency joint analysis. It can only simply convert time-domain signals into frequency-domain signals, but cannot obtain the time when specific frequency band data appears, which has serious drawbacks in data processing; HHT When the transformation decomposes complex signals, it has the disadvantages of low accuracy of solution results and long calculation time. At the same time, there are problems of edge effects, cross-border problems, stopping criteria and filtering of false low-frequency components.

The short-time Fourier transform has a better effect on the set frequency when extracting the envelope, and the signal processing effect on other frequencies is poor.

Wavelet transform can divide data into n segments for processing, but simple wavelet transform is not intelligent enough and requires professionals to identify signals. In ultrasonic nondestructive testing, noise is often mixed with useful signals, so the key to obtaining defect information is to noise reduction

Images