基于声信号的供水管道泄漏检测定位系统研究

发布时间：2018-02-24 23:34

本文关键词： 管道泄漏声信号时延估计变步长LMS 遗忘加权　出处：《辽宁工程技术大学》2015年硕士论文　论文类型：学位论文

【摘要】：城市的供水管道线路复杂,分布范围广,运输距离长。在自然老化、腐蚀等非人为因素条件下,管道泄漏事故经常发生。然而传统人工寻找泄漏点的方式效率低下,往往因为对泄漏事故处理不及时而造成严重的资源浪费和经济损失。因此,如何快速的查找管道泄漏点,实现泄漏的在线检测和定位已成为当下一个非常紧迫的课题。对现有的管道泄漏检测方法进行分析对比,根据现场管道实际运行环境,采用了基于声信号的管道泄漏检测方式。针对传统基于双传感器的管道泄漏检测定位模型中声速采用经验值固定输入,导致定位精度低的问题,将声速固定输入的方式改进为由现场实时检测的方式,构建了基于三传感器的管道泄漏检测定位模型。研究了不同管道材质、管道压力、声信号传播距离、泄漏点口径以及管道接口材质五种工况条件下的泄漏声信号时频特性的变化规律,对声信号在管道传播中的衰减特性做了实验分析,了解管道泄漏声信号的传播特性。分析现有的变步长最小均方LMS自适应时延估计算法的原理,针对步长因子在算法收敛程度加深时快速变小的问题,对原有的步长因子迭代函数进行改进,引入历史误差平方的遗忘加权和补偿项,并加入滑动窗,提高算法的时延估计性能。设计了系统的各部分硬件,包括信号调理电路、AD转换模块、GPS授时模块、GPRS数据传输模块、电源模块以及DSP最小系统等。采用VC++和MATLAB混合编程的方式完成上位机软件的开发。本文以基于三传感器的供水管网漏点检测定位模型为硬件实验平台,结合上位机软件对管道的泄漏检测和定位效果进行测试,泄漏检测定位测试结果表明该系统能够正确的检测到泄漏信号,并且对泄漏点定位精度较高,满足实际生产的需要。
[Abstract]:Urban water supply pipelines are complex, wide distribution, long transportation distance. Pipeline leakage accidents often occur under the condition of natural aging, corrosion and other non-human factors. However, the traditional artificial way of finding leak points is inefficient. It often causes serious waste of resources and economic loss because the leak accident is not dealt with in time. Therefore, how to find the leak point of pipeline quickly, On-line detection and location of leakage has become a very urgent issue. The existing pipeline leak detection methods are analyzed and compared, according to the actual operation environment of pipeline, The method of pipeline leakage detection based on acoustic signal is adopted. In the traditional model of pipeline leakage detection based on two sensors, the speed of sound is fixed by the experience value, which leads to the problem of low positioning accuracy. In this paper, the fixed input mode of sound velocity is improved to the real-time detection mode in the field, and a three-sensor model for pipeline leakage detection is constructed. The different pipeline materials, pipeline pressure, acoustic signal propagation distance are studied. The variation law of time-frequency characteristic of leakage acoustic signal under five working conditions of leak caliber and pipeline interface material is analyzed experimentally, and the attenuation characteristic of acoustic signal in pipeline propagation is analyzed experimentally. This paper analyzes the principle of the existing adaptive time delay estimation algorithm of variable step minimum mean square (LMS), aiming at the problem that step factor decreases rapidly when the convergence degree of the algorithm deepens. The original step size factor iteration function is improved, and the forgetting weight and compensation term of the square of historical error is introduced, and sliding window is added to improve the time delay estimation performance of the algorithm. The hardware of each part of the system is designed. It includes signal conditioning circuit and AD conversion module, GPS timing module and GPRS data transmission module. The power module and the DSP minimum system are used to develop the upper computer software by the mixed programming of VC and MATLAB. In this paper, the leak detection and location model of water supply network based on three sensors is used as the hardware experimental platform. Combined with the software of upper computer, the leakage detection and location effect of pipeline is tested. The result of leak detection and location test shows that the system can detect leakage signal correctly, and the accuracy of leak location is high, which meets the needs of actual production.
【学位授予单位】：辽宁工程技术大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TU991.33

【参考文献】