用自适应滤波解卷积进行缺陷类型识别
Classification of flaw by adaptive filtering deconvolution of backscattering ultrasonic echo
-
摘要: 提出了超声缺陷散射回波信号的卷积模型和缺陷散射回波信号的自适应滤波解卷积方法。对一给定系统的模拟,进行了解卷积效果分析.通过解卷积,能够减弱其它因素的影响,突出缺陷的类型(形状)和大小信息。通过对横穿孔、平底孔、球孔和长方孔缺陷的共18个样品的回波信号进行解卷积处理,实现了带棱边的平面与光滑曲面──自然缺陷二要素的类型识别。Abstract: A convolution model of flaw scattering echoes, and an adaptive deconvolution method are presented. The effect of the method is analyzed by simulating a given system. By deconvolution, the influence of the transducer is eliminated greatly, and the flaw characteristic information stand out moreclearly in the deconvolved echoes than in flaw echoes itself. Flaw echo signals of 18 practical samples are processed with the method. Flaws are classified successfully.