摘要
为了及时检测输油管道泄漏缺陷以减少输油泄漏事故的发生,首先,提出了一种基于小波包(WPT)能量谱和希尔伯特变换(Hilbert)的输油管道泄漏检测技术,改变阀门开度模拟不同泄漏程度的管道泄漏;其次,选取小波包阈值对采集的信号去噪,获得实际有效信号,并用WPT能量谱变换将有效信号分解为频段能量谱,过滤掉不含泄漏信息的频段,保留有效频段;最后,利用Hilbert进行有效频段的谱分析,提取泄漏源信息,实现对管道泄漏工况的诊断分类。结果表明:输油管道泄漏检测平均误差在3%以内,诊断准确率达到95%,与直接Hilbert变换方法以及单纯WPT能量谱分析法相比,定位精度明显提高。
In order to detect pipeline leakage defects in time to reduce the occurrence of oil transmission leakage ac cidents,firstly,a pipeline leakage detection technique based on wavelet packet(WPT)energy spectrum and Hilbert transform is proposed to simulate pipeline leakage of different leakage degrees by changing the valve opening degree;then,select wavelet packet threshold to de-noise the collected signal to obtain the actual effective signal,and use WPT energy spectrum transformation to decompose the effective signal into the frequency band energy spectrum,filter out the frequency band without leaking information,and keep effected signal.Finally,Hilbert is used to perform spectral analysis of effective frequency bands to extract leakage source information,so as to realize the diagnosis and classifi cation of pipeline leakage conditions.The results show that the average error of oil pipeline leak detection is within 3%,and the diagnostic accuracy rate reaches 95%.Compared with the direct Hilbert transform method and the simple WPT energy spectrum analysis method,the positioning accuracy is significantly improved.
作者
王新颖
林振源
冯珊珊
胡磊磊
徐拓
刘岚
陈海群
WANG Xinying;LIN Zhenyuan;FENG Shanshan;HU Leilei;XU Tuo;LIU Lan;CHEN Haiqun(College of Environmental and Safety Engineering,Changzhou University,Changzhou Jiangsu 213164,China)
出处
《工业安全与环保》
2023年第6期22-27,共6页
Industrial Safety and Environmental Protection
基金
常州市国际科技合作项目(CZ20210026)。
作者简介
王新颖(1976-),女,硕士,副教授,主要从事安全检测、监控方面的研究。E-mail:wangxy@cczu.edu.cn。
