摘要
首先选择节点压力组成燃气管网泄漏工况特征向量,研究分析发现其在实际应用中具有一定的局限性,为此进一步提出了管网泄漏工况特征图形化的方法对其进行优化。利用Pipeline Studio建立了经过实验验证的计算机模拟管网,通过稳态模拟获取数据进而绘制燃气管网泄漏工况特征图像并进行对比分析,最终论证了燃气管网泄漏工况特征图形在基于模式识别的燃气管网泄漏检测定位中应用的可行性。
Firstly the nodes pressure is selected to comprise the gas pipeline networks operating condition feature vector, it is found that the feature vector based on nodes pressure have some limitations in practical applications, and so the feature graphical method is put forward to optimize the feature vector. In the research, the Pipeline Studio is applied to set up an experimentally validated computer simulation pipeline network and the gas pipeline networks operating condition feature images are drawn according to the steady - state simulation data. Finally it is demonstrated by the comparative analysis of the feature images that the feasibility of using gas pipeline networks operating condition feature images in leak detection and location based on pattern recognition.
作者
靳新迪
玉建军
李宏阳
JIN Xindi YU Jianjun LI Hongyang(Tianjin Chengjian University Tianjin 30038)
出处
《工业安全与环保》
北大核心
2017年第3期43-47,共5页
Industrial Safety and Environmental Protection
关键词
燃气管网
特征选择
泄漏检测和定位
模式识别
图像处理
gas pipeline network feature selection leak detection and location pattern recognition imagery processing