光学 精密工程, 2019, 27 (4): 932, 网络出版: 2019-07-25
基于超声背散射信号递归分析的金属材料微缺陷识别
Micro defects detection in metallic materials based on recurrence analysis of ultrasonic backscattering signal
金属材料 背散射信号 递归分析 递归图 递归定量分析 metallic material backscattered signal recurrence analysis recurrence plot recurrence quantification analysis
摘要
为了能有效识别金属材料超声检测信号中的微小缺陷回波, 使用递归分析方法对检测信号进行分析。通过对超声波背散射信号进行建模, 说明其组成要素。在超声检测信号中, 缺陷回波信号会对系统的递归特性产生影响。使用递归分析对含0.8 mm平底孔人工模拟缺陷的直径为120 mm低碳钢棒材试块实验采集的背散射信号和无缺陷背散射信号进行了研究。截取试块检测信号中的背散射信号部分, 通过合理的参数选择对其进行递归分析并绘制递归图。通过与实验采集的无缺陷信号的递归图进行对比, 发现缺陷信号会在递归图中产生明显的白色交叉条纹带。使用递归定量分析进一步研究了含缺陷背散射信号的递归特征量, 结果表明捕获时间(TT)、确定率(DET)与递归熵(ENTR)这三项特征量对缺陷信号比较敏感, 在缺陷位置处均会出现明显的峰值。
Abstract
To recognize minor defect echoes from metallic material ultrasonic testing signals, the recurrence analysis method was used to analyze ultrasonic detecting signals. By modeling the backscattering signal, we describe its components. In the ultrasonic detection signal, the defect echo signal will affect the recurrence characteristics of the system. In this study, we employ the recurrence analysis method to study the backscattering defect and defect-free signals of a low carbon steel testing specimen with a diameter of 120 mm. This specimen was artificially simulated with a 0.8-mm flat bottom hole. Through reasonable parameter selection, we performed recurrence analyses of the backscattering signals intercepted from the detection signals and produced recurrence plots. A comparison with the recurrence plot of non-defective signals showed that the defect signal will produce an obvious white cross stripe in its recurrence plot. We further studied the recurrence characteristic quantities of the backscattering signals with defects by conducting a recurrence quantification analysis. The results show that the three characteristic quantities, including trapping time, determinism, and entropy, are more sensitive to the defect signals, and all exhibit obvious peaks at the defect location.
杨辰龙, 冯玮, 边成亮, 周晓军, 柴景云. 基于超声背散射信号递归分析的金属材料微缺陷识别[J]. 光学 精密工程, 2019, 27(4): 932. YANG Chen-long, FENG Wei, BIAN Cheng-liang, ZHOU Xiao-jun, CHAI Jing-yun. Micro defects detection in metallic materials based on recurrence analysis of ultrasonic backscattering signal[J]. Optics and Precision Engineering, 2019, 27(4): 932.