易方, 李著信, 苏毅, 王鹏飞, 吴昊. 基于改进型小波阈值的输油管道磁记忆信号降噪方法[J]. 石油学报, 2009, 30(1): 141-144. DOI: 10.7623/syxb200901032
引用本文: 易方, 李著信, 苏毅, 王鹏飞, 吴昊. 基于改进型小波阈值的输油管道磁记忆信号降噪方法[J]. 石油学报, 2009, 30(1): 141-144. DOI: 10.7623/syxb200901032
YI Fang, LI Zhuxin, SU Yi, WANG Pengfei, WU Hao. Denoising algorithm for metal magnetic memory signals of oil pipeline based on improved wavelet threshold[J]. Acta Petrolei Sinica, 2009, 30(1): 141-144. DOI: 10.7623/syxb200901032
Citation: YI Fang, LI Zhuxin, SU Yi, WANG Pengfei, WU Hao. Denoising algorithm for metal magnetic memory signals of oil pipeline based on improved wavelet threshold[J]. Acta Petrolei Sinica, 2009, 30(1): 141-144. DOI: 10.7623/syxb200901032

基于改进型小波阈值的输油管道磁记忆信号降噪方法

Denoising algorithm for metal magnetic memory signals of oil pipeline based on improved wavelet threshold

  • 摘要: 利用金属磁记忆检测方法对输油管道进行早期诊断时,磁记忆检测信号常常被各种噪声源污染,极大地降低缺陷信号可检测性。在传统软、硬阈值降噪方法基础上,根据磁记忆检测信号的特点,提出了一种改进小波阈值函数与自适应阈值相结合的方法,选用Daubechies小波作为小波函数,分解级数为4层,采用自适应方法计算阈值。用新型漏磁/磁记忆检测仪进行了算法验证,将改进降噪方法应用于磁记忆检测信号的降噪处理。与传统软、硬阈值降噪算法相比,新算法克服了软阈值信号失真和硬阈值不连续、振荡等缺点,提高了重建信号的信噪比,降低了均方根误差值,有效地消除了信号噪声,为正确判断输油管道的应力集中位置及早期诊断提供了理论依据。

     

    Abstract: When the metal magnetic memory(MMM) inspection method is applied to make early diagnosis of oil pipeline, the magnetic memory signal is easy to be disturbed by various sources of noises, and the detectability of defect signals is greatly lowered. An improved method composed of wavelet threshold function and adaptive threshold was presented on the basis of the classic denoising method using soft and hard threshold algorithm and the characteristics of magnetic memory signals. The Daubechies wavelet was used as a wavelet function with four series, and the adaptive method was used to calculate the threshold. The magnetic flux leakage-magnetic memory detector was used to certify the algorithm. The testing results demonstrated that the new method could overcome the shortcomings of soft and hard threshold algorithm, and the signal-to-noise ratio of the rebuilt signal was improved, and the noises of signal were greatly eliminated. This denoising algorithm can provide the theoretical basis for estimation of stress concentration zone and early diagnosis of oil pipeline.

     

/

返回文章
返回