石油学报 ›› 2014, Vol. 35 ›› Issue (6): 1197-1203.DOI: 10.7623/syxb201406018

• 石油工程 • 上一篇    下一篇

管道泄漏信号和干扰信号的数字化判别方法

林伟国, 王晓东, 戚元华, 吴海燕   

  1. 北京化工大学信息科学与技术学院 北京 100029
  • 收稿日期:2014-04-12 修回日期:2014-07-27 出版日期:2014-11-25 发布日期:2014-10-13
  • 通讯作者: 吴海燕,女,1980年8月生,2001年获北京化工大学自动化专业学士学位,2012年获北京化工大学控制理论与控制工程专业博士学位,现为北京化工大学信息科学与技术学院讲师,主要从事生产过程智 能检测与安全预警技术方面的研究工作。Email:wuhy@mail.buct.edu.cn
  • 作者简介:林伟国,男,1968年2月生,1990年获中国矿业大学自动化专业学士学位,1999年获哈尔滨工业大学精密仪器与机械专业博士学位,现为北京化工大学信息科学与技术学院教授,主要从事生产过程智能检测与安全预警技术方面的研究工作。Email:linwg@mail.buct.edu.cn
  • 基金资助:

    国民核生化灾害防护国家重点实验室开放基金资助项目(SKLNBC2014-10)资助。

Digital identification method for pipeline leak and interference signal

Lin Weiguo, Wang Xiaodong, Qi Yuanhua, Wu Haiyan   

  1. College of Information Science and Technology, Beijing University of Chemical Technology, Beijing, 100029
  • Received:2014-04-12 Revised:2014-07-27 Online:2014-11-25 Published:2014-10-13

摘要:

为了降低管道泄漏检测过程中调压、调泵和阀门动作等干扰信号引起的泄漏诊断误报率,在理论分析干扰信号互相关判别法存在问题的基础上,提出基于短间距安装双传感器结构的泄漏信号和干扰信号的数字化判别算法。把干扰信号和泄漏信号都归为异常信号,通过小波去噪、异常信号识别及异常信号提取等处理,提取出两路传感器输出信号中的异常信号跳变沿并计算两路信号的累积差分,根据累积差分的符号判断出双传感器信号的时移方向,进而判断出异常信号来源,实现干扰信号与泄漏信号的甄别。该方法大大降低了双传感器结构对安装间距的要求,克服了双传感器结构的硬件电路延时法对传感器灵敏度、响应特性和信号放大倍数的苛刻要求,在较低信号采样率的条件下实现了干扰信号的准确识别。实验室模拟异常信号的处理结果表明,对于液态介质(水),该方法可以实现干扰和泄漏信号的准确判别,双传感器安 装间距可以缩小到1.5 m以内,采样频率可以降低到50 Hz。

关键词: 管道泄漏检测, 双传感器结构, 短间距安装, 异常信号, 数字化识别

Abstract:

Interference signals related to the actions of regulator, pump and valve often cause false alarms in pipeline leak detection. A digital method for identifying pipeline leak and interference signals based on dual sensors installed at a short distance is proposed after theoretical analysis of the existing problems in interference identification by cross-correlation. In the proposed algorithm, both interference and leak signals are classified as abnormal signals. Through wavelet de-noising and abnormal signal recognition and extraction, the rising edge of abnormal signals from the dual sensors is extracted and cumulative difference of the two sets of signals is then calculated. The time shift direction as well as the source of signals from the dual sensors is judged, and the recognition of interference and leak signals is further implemented. This method greatly reduces the installation distance of dual sensors, overcomes the strict requirements for consistence of sensor sensitivity, response characteristics, and signal amplification in hardware circuit delay method, and achieves accurate recognition of interference signals at low sampling frequency. Laboratory simulation of abnormal signals proved that using using the proposed method, the interference can be identified accurately in aqueous medium (water) with the installation distance of dual sensors reduced down to 1.5 m and the sampling frequency to 50 Hz.

Key words: pipeline leak detection, dual sensor structure, short-distance installation, abnormal signal, digital identification

中图分类号: