| 基于声信号特征分析的燃气管道探测识别方法* |
| |
| 引用本文： | 刘恩斌,温櫂荣,郭冰燕,喻斌,陈其琨.基于声信号特征分析的燃气管道探测识别方法*[J].中国安全生产科学技术,2022,18(4):61-68. |
| |
| 作者姓名： | 刘恩斌 温櫂荣 郭冰燕 喻斌 陈其琨 |
| |
| 作者单位： | （1.西南石油大学 石油与天然气工程学院，四川 成都 610500；2.中国石油化工股份有限公司 天然气分公司，北京 100029；3.中石油华北油田分公司，河北 任丘 062550；4.中国石油管道局工程有限公司，河北 廊坊 065000；5.School of Engineering,Cardiff University，UK Cardiff CF24 3AA） |
| |
| 基金项目： | * 基金项目: 中石油重大科技项目（2021DJ2804）；四川省应用基础项目（2019YJ0352） |
| |
| 摘 要： | 为了探测和辨识地下燃气管道，避免燃气管道改扩建的过程发生第三方破坏引发安全事故，提出1种基于声信号特征分析的燃气管道探测识别方法，该方法考虑燃气管道声信号声压级低以及易衰减的特点，采用Hilebert-Huang变换算法分析燃气管道流噪声信号特征，建立燃气管道流噪声信号的特征数据库，并通过BP神经网络进行模式识别，判别管道的种类以及在役状态。通过对实测数据和数值模拟数据的分析表明:该方法的有效识别率达到了97.5％，验证了该方法的有效性。
|
| 关 键 词： | 燃气管道 声信号 希尔伯特黄变换 特征提取 BP神经网络 模式识别 |
Detection and recognition methods of gas pipelines based on acoustic signal feature analysis |
| |
| LIU Enbin,WEN Zhaorong,GUO Bingyan,YU Bin,CHEN Qikun.Detection and recognition methods of gas pipelines based on acoustic signal feature analysis[J].Journal of Safety Science and Technology,2022,18(4):61-68. |
| |
| Authors: | LIU Enbin WEN Zhaorong GUO Bingyan YU Bin CHEN Qikun |
| |
| Affiliation: | (1.Petroleum Engineering School,Southwest Petroleum University,Chengdu Sichuan 610500,China;2.SINOPEC Gas Company,Beijing 100029,China;3.CNPC Huabei Oil Field Branch,Renqiu Hebei 062550,China;4.China Petroleum Pipeline Engineering Corporation,Langfang Hebei 065000,China;5.School of Engineering,Cardiff University,Cardiff CF24 3AA,UK) |
| |
| Abstract: | In order to detect and identify underground gas pipeline and to avoid safety accidents caused by third-party damage during the reconstruction and expansion of gas pipeline,a gas pipeline detection and identification method based on acoustic signal feature analysis was proposed，considering the characteristics of low sound pressure level and easy attenuation of gas pipeline acoustic signal.The Hilbert Huang transform algorithm was used to analyze the characteristics of gas pipeline flow noise signal,and the characteristic database of gas pipeline flow noise signal was established,BP neural network was used for pattern recognition to distinguish the type of pipeline and in-service state.The analysis of measured data and numerical simulation data showed that the effective recognition rate of this method reaches 97.5%,which verifies the effectiveness of this method. |
| |
| Keywords: | gas pipeline acoustic signal Hilbert-Huang transform feature extraction BP neural network pattern recognition |
|
| 点击此处可从《中国安全生产科学技术》浏览原始摘要信息 |
|
点击此处可从《中国安全生产科学技术》下载全文 |
