测量矩阵对管道超声压缩采样系统性能的影响  

Effect of measurement matrix to the performance of pipeline ultrasonic compressive sampling system

作　　者：王云蛟[1] 宋寿鹏[1] 堵莹 邵勇华 彭成庆 

机构地区：[1]江苏大学仪器科学与工程系,江苏镇江212013

出　　处：《信息技术》2015年第11期21-24,33,共5页Information Technology

基　　金：国家自然科学基金资助项目(51375217)

摘　　要：基于压缩传感(Compressed Sensing,CS)理论的管道超声压缩采样系统可用于解决长距离能源输送管道在役超声检测数据量大的问题,然而该系统的性能受到多种因素的影响,测量矩阵的多样性是最重要的影响因素之一。针对此问题,采用不同于约束等距性(Restricted Isometric Property,RIP)的统计方法分析了测量矩阵对系统性能的影响机理,提出了优质原子比和相对渡越时间误差作为衡量测量矩阵性能优劣的指标。管道超声检测实验结果表明,该项技术可作为压缩采样中测量矩阵的选择依据,为管道中进行超声检测信号压缩采样提供借鉴。Pipeline ultrasonic compressive sampling system based on compressed sensing （CS）theory can be used to solve the problem of huge amount of data acquired from ultrasonic detection in long distance transportation pipeline. However, the performance of the system is affected by many factors and the diversity of measurement matrix is one of the most important ones. To solve this problem, this paper analyzes the influence mechanism of measurement matrix on the system performance using the statistical method different from the restricted isometric property（RIP） and proposes the ratio of high quality atoms and relative error of transit time as indicators to measure the merits of the measurement matrix. The ultrasonic pipeline detection experimental results show that this technology can be used as selection criteria of measurement matrix and provide a reference for pipeline ultrasonic compressive sampling.

关 键 词：管道 超声检测 压缩采样 测量矩阵 

分 类 号：TP274.53[自动化与计算机技术—检测技术与自动化装置]