基于全频带小波能量相对熵的脉冲涡流感应信号辨识  

Recognizing the Pulsed Eddy Current Inductive Signal Based on Full Waveband Relative Entropy of Wavelet Energy

作　　者：杨正理[1] 陈海霞[1] 徐智[1] 

机构地区：[1]三江学院机械与电气工程学院,江苏南京210012

出　　处：《自动化仪表》2017年第9期75-80,共6页Process Automation Instrumentation

摘　　要：脉冲涡流检测是一种新兴的无损检测技术,检测方便、成本低、安全性好,适用于对各种材料缺陷进行在线检测。然而,脉冲涡流感应信号强度微弱、识别困难,必须采用比较有效的信号处理方法对信号进行处理。为了提高脉冲涡流检测系统的精度和正确率,在详细分析了脉冲涡流感应信号特性后,提出了一种基于全频带小波能量相对熵的脉冲涡流感应信号辨识方法。对几种典型脉冲涡流感应信号的采样数据进行多尺度小波分解,并单支重构各尺度小波系数。利用小波变换对局部信号特征的放大作用以及小波相对熵对信号之间差异的辨识优势,计算各尺度小波能量占所有小波系数能量的权重系数,及原始信号在各频带下的小波能量相对熵,然后根据小波能量相对熵对脉冲涡流感应信号的类别进行信号辨识。理论分析和试验表明,该方法能够有效消除干扰,提高无损检测的精度和正确率。Pulsed - eddy current detection is a new type of nondestructive detection technology, which features convenient detection,low cost and safety,so it is suitable for various kinds of online detection of material defects. However,the intensity of pulsed - eddy current inductive signal is so weak that it is difficult to identify the signals, and more effective signal processing method must be adopted to deal with the weak signals. In order to improve the precision and accuracy of pulsed - eddy current testing system, the characteristics of pulsed - eddy current inductive signal are analyzed in detail, then a new identification method of pulsed - eddy current inductive signal based on the full waveband relative entropy of wavelet energy is proposed. The multi - scale wavelet decomposition is carried out for several typical sample data of the pulsed - eddy current inductive signals, and the wavelet coefficient of each scale is reconstructed. By using the amplification function of wavelet transform to local signal characteristics and the identification superior of wavelet relative entropy to the difference among signals, the weight coefficients of each scale wavelet energy in all the wavelet coefficient energy are calculated, and each signal wavelet energy relative entropy in each frequency band is calculated, then according to these characteristics of wavelet energy relative entropy, the category of pulsed - eddy current inductive signal is identified. Theoretical analysis and experimental results show that such method effectively avoids disturbance, and improves the precision and accuracy of nondestructive detection.

关 键 词：小波能量相对熵 脉冲涡流感应信号 信号辨识 无损检测 小波变换 多尺度分解 

分 类 号：TH-39[机械工程] TP73[自动化与计算机技术—检测技术与自动化装置]