电磁声发射信号的压缩与重构处理

Compression and Reconstruction Processing of the Electromagnetically Induced Acoustic Emission Signal

电磁声发射是一种新型的无损检测技术,其通过对金属构件进行电磁加载,在缺陷处产生声发射信号,根据信号特征对金属构件进行缺陷检测。针对该技术在测定缺陷位置和类型时连续多次加载而产生的大量数据问题,引入信号的压缩感知理论,基于正交匹配追踪算法分别对两种不同类型的声发射信号和电磁声发射信号进行压缩重构,选取不同的测量值,研究其对信号重构效率的影响,从波形和频谱两方面分析重构效果。实验结果表明:压缩感知理论能够达到压缩电磁声发射信号的目的,并且测量值和重构误差之间呈指数衰减关系,与重构时间呈线性增长关系,综合考虑压缩重构的各方面因素,当测量值取8~10倍的信号稀疏度时,电磁声发射信号能够获得较高的重构效率。

Electromagnetically induced acoustic emission（ EMAE） is a new technique of nondestructive testing,it can electromagnetically load to the metal components and generate the acoustic emission signal,which can be used to test metal components according to the result of signal processing. For acquiring the location and size of the defect,it needs to repeatedly load which will generate a lot of data. The theory of compressed sensing was introduced to compress and reconstruct the acoustic emission（ AE） and EMAE signals based on orthogonal matching pursuit（ OMP）. The different measurement numbers were selected to contrast the result on signal reconstruction. The reconstruction effect was analyzed on the waveform and spectrum. Experimental results show that it can be achieved to compress the EMAE signals by using the method of compressed sensing and the measured number is exponential decaying with the reconstruction error and in direct proportion to the reconstruction time. Based on an overall consideration of various factors,the EMAE signal can obtain higher efficiency of reconstruction when the measured number is 8 ~ 10 times of the signal sparseness.