摘要
管道泄漏通常会造成管体的变形和破坏,利用金属磁记忆技术可有效检测由于泄漏而造成的管道应力集中。针对含有干扰噪声的非稳态磁记忆信号泄漏点特征信息提取,提出一种基于希尔伯特–黄变换的信号分析方法。该方法首先利用经验模态分解(Empirical Mode Decomposition,EMD)将磁记忆信号分解成本征模式函数(Intrinsic Mode Function,IMF(Mi(t))),分离和提取Mi(t)分量并对信号进行重构,然后通过对重构信号的Hilbert包络谱进行分析,可以达到提取管道泄漏点特征信息的目的。实验结果表明,该方法不受管道压力、泄漏量等参数的影响,能够有效提取管道泄漏点的特征信息,具有很好的泄漏点识别准确率。现场检测结果表明,利用该方法对油气管道泄漏点的定位误差小于±1.0 m,验证了该方法的有效性。
Underground metal pipeline leakage detection and positioning is an urgent technical problem. Usually,the leakage could cause the pipe wall to be deform and damage. The metal magnetic memory is an effectively method to detect the stress concentration. In order to extract the feature of pipeline leakage from the magnetic signals,a time-frequency analysis method has been proposed based on Hilbert-Huang Transform. Firstly,the Intrinsic Mode Functions（IMF,Mi（t））of magnetic signals were obtained using the empirical mode decomposed（EMD）algorithm. Then,through the separation and extraction of the different frequency components,the reconstructed signal by low-frequency Mi（t） would contain feature of pipeline leakage. Finally,the purpose has been realized to extract the feature of pipeline leakage according to analysis of the Hilbert envelope spectrum. The favorable recognition precision ratio and the validity of extraction feature of pipeline leakage were verified by the experiments. Furthermore,experimental results indicate that the pressure and leakage could not affect the extraction feature. The results of in-situ show that the accuracy of leakage positioning is less than±1.0 m.
出处
《西南石油大学学报（自然科学版）》
CAS
CSCD
北大核心
2013年第5期165-171,共7页
Journal of Southwest Petroleum University(Science & Technology Edition)
关键词
金属磁记忆
希尔伯特-黄变换
埋地金属管道
泄漏点检测
定位
metal magnetic memory
Hilbert-Huang Transform（HHT）
underground metal pipeline
leakage detection
localization
