• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2018, Vol. 54 ›› Issue (18): 33-41.doi: 10.3901/JME.2018.18.033

• 仪器科学与技术 • 上一篇    下一篇

基于贪婪算法的汽轮机叶轮相控阵信号压缩感知

白志亮, 陈世利, 贾乐成, 曾周末   

  1. 天津大学精密测试技术与仪器国家重点实验室 天津 300072
  • 收稿日期:2017-09-07 修回日期:2017-11-06 出版日期:2018-09-20 发布日期:2018-09-20
  • 通讯作者: 陈世利(通信作者),男,1973年出生,博士,副教授。主要研究方向为无损检测技术、管道泄漏智能检测。E-mail:shlchen@tju.edu.cn
  • 作者简介:白志亮,男,1992年出生,博士研究生。主要研究方向为超声相控阵检测、压缩感知算法。E-mail:zhl_bai@tju.edu.cn
  • 基金资助:
    科技部国家重点研发计划资助项目(2016YFF0101800)。

Compressed Sensing of Ultrasonic Phased Array Signal in Turbine Disc Rims Inspection Based on Greedy Algorithms

BAI Zhiliang, CHEN Shili, JIA Lecheng, ZENG Zhoumo   

  1. State Key Laboratory of Precision Measurement Technology and Instrument, Tianjin University, Tianjin 300072
  • Received:2017-09-07 Revised:2017-11-06 Online:2018-09-20 Published:2018-09-20

摘要: 超声相控阵是核电站组件无损检测中广泛采用的手段之一。针对低压汽轮机叶轮轮缘缺陷检测中存在的数据量大的问题,提出一种基于贪婪算法的超声相控阵信号压缩感知方法。利用CIVA平台建立了超声相控阵缺陷检测仿真模型,使用四种贪婪算法对仿真信号进行压缩感知并重构,计算不同采样率和不同压缩率下的百分比均方误差,根据结果选取最优算法;使用上述算法对汽轮机叶轮模型电火花加工缺陷回波信号进行压缩重构;通过与小波压缩重构精度的对比,验证该算法在超声相控阵汽轮机叶轮缺陷检测中的适用性。结果表明,使用仿真数据可以用低于奈奎斯特极限的测量点数精确重构信号;在压缩率为60%时,使用试验信号的平均重构误差仅为4.815 2%,与小波压缩的重构精度相当。

关键词: 超声相控阵, 汽轮机叶轮, 贪婪算法, 无损检测, 压缩感知

Abstract: Ultrasonic phased array is one of the most widely used approaches in the nuclear power plants nondestructive testing (NDT). Aiming at reducing the amount of data processed in defect detection of the turbine disc rims, a compressed sensing (CS) of ultrasonic phased array signals is proposed based on greedy algorithms. Firstly, a simulation model of ultrasonic phased array defect detection is established using CIVA platform. Considering the influence of sampling rate, the optimal algorithm is selected among four greedy algorithms according to the percent residual difference (PRD) in different compression ratio (CR) levels. The reconstruction performance is then experimentally evaluated using the echo signals extracted from EDM notches of artificial mockup. By the comparison with conventional discrete wavelet transform (DWT) compression, the practicability of proposed CS applied for phased array signals in defect detection of the turbine disc rims is finally analyzed. The simulated results demonstrate that excellent recoveries can be obtained using measurement points less than the Nyquist sampling limitation required. Besides, the experimental performances of CS are comparable with that of DWT compression with PRD reaching 4.8152% when CR is 60%.

Key words: compressed sensing, greedy algorithms, nondestructive testing, turbine disc rim, ultrasonic phased array

中图分类号: