摘要
首先分析管道泄漏检测的基本原理 ,并应用LabVIEW分析单传感器在泄漏管道不同位置拾取的泄漏信号的时频域特征 ,以所提取时频域特征指标来构造人工神经网络的输入矩阵 ,建立了能对管道泄漏状况进行分析检测定位的人工神经网络 ,实现了管道泄漏检测的单传感器定位。模拟泄漏实验和在PC机上进行网络仿真表明该方法是有效的。
The principle of leakage detection of pipelines was introduced, and the characteristics of the leakage signals in time and frequency domains collected with single sensor in different positions of pipelines were analyzed with LabVIEW. The artificial neural network (ANN) with the input matrix composed of the characters of stress was established to detect and localize the leakage in pipelines. Simulating experiment showed the effectiveness of the method.
出处
《无损检测》
2002年第8期328-331,共4页
Nondestructive Testing
