Wavelet decomposition and extraction method of defect magnetic flux leakage testing signals based on waveform similarity
YANG Jie1, LI Hongmei2, ZHAO Chuntian1,2, YANG Hongli3
1. Southern University of Science and Technology, Shenzhen, Guangdong 518055; 2. Sichuan University, Chengdu 610207; 3. Shandong University of Science and Technology, Qingdao, Shandong 266590
Abstract：Magnetic flux leakage testing (MFLT) technology has been widely used in engineering to identify and measure defects in ferromagnetic components. Its capability in reconstructing defects in the three-dimensional (3D) form, especially for general complex defects formed naturally (such as by corrosion and/or fatigue), is one of the primary indexes representing the technological advances. Because the actual defect shape is irregular and often contains multiple pits, its magnetic flux leakage detection signals affect each other. To improve the accuracy of reconstructing defects in 3D and speed up the data processing, the prerequisite is to automatically strip off and extract the valid defect data from the MFLT signals. To effectively locate, separate, and integrate information for reconstructing each defect, a wavelet decomposition and extraction method is proposed. Based on the morphological characteristics of the MFLT signals, a wavelet basis is selected that matches the characteristics of the defect signals closely. This wavelet basis is used in the multi-scale decomposing of the MFLT signals and in the calculating of the associated wavelet high frequency coefficients. Then the wavelet coefficient is utilized in identifying and locating the defects in each channel of the signals, as well as in extracting the MFLT data corresponding to each defect. Finally, by examining the signals of each defect in a channel and integrating all the associated data in the adjacent channels, a complete set of data related to each defect is obtained. The proposed method can be applied to quickly identify, extract and integrate the MFLT data for all defects, which provides the basis for the 3D defect reconstruction.
杨杰, 李红梅, 赵春田, 杨洪礼. 基于相似波形的漏磁检测缺陷信波分解及提取方法[J]. 电气技术, 2022, 23(6): 8-16.
YANG Jie, LI Hongmei, ZHAO Chuntian, YANG Hongli. Wavelet decomposition and extraction method of defect magnetic flux leakage testing signals based on waveform similarity. Electrical Engineering, 2022, 23(6): 8-16.
2022, Vol. 23 
