【摘要】：保障管道安全生產(chǎn)有諸多技術(shù)手段,其中最重要并被廣泛應(yīng)用的是泄漏檢測技術(shù)。管道在運(yùn)行維護(hù)時(shí),需要對(duì)管道信號(hào)進(jìn)行實(shí)時(shí)監(jiān)測,準(zhǔn)確判斷管道運(yùn)行狀況,及時(shí)診斷出存在的泄漏隱患,避免安全事故的發(fā)生。天然氣管道泄漏產(chǎn)生的次聲波信號(hào)屬于非平穩(wěn)信號(hào),且在管道傳播過程中受到嚴(yán)重的外界干擾,導(dǎo)致泄漏的誤判率高,出現(xiàn)錯(cuò)報(bào)、漏報(bào)、誤報(bào)的情況。自適應(yīng)時(shí)頻分析方法是分析非平穩(wěn)信號(hào)的重要手段,針對(duì)實(shí)驗(yàn)室采集的管道泄漏產(chǎn)生的次聲波信號(hào)的特征,本文利用基于相關(guān)系數(shù)的變分模態(tài)分解(Variational Mode Decomposition,簡稱VMD)算法選取預(yù)設(shè)尺度K值,再通過基于互信息的VMD算法進(jìn)行管道特征提取,從而進(jìn)行工況識(shí)別,從而提高系統(tǒng)對(duì)管道泄漏判斷的準(zhǔn)確率。本文主要工作有:1、查閱有關(guān)管道泄漏檢測相關(guān)論文,深入了解管道泄漏檢測的相關(guān)方法,選取VMD方法應(yīng)用到天然氣管道泄漏檢測中;對(duì)管道泄漏檢測存在的故障情況進(jìn)行分析,對(duì)故障情況進(jìn)行分類,最終確定模式識(shí)別的目標(biāo)種類;完成不同管道運(yùn)行情況下的信號(hào)采集。2、針對(duì)VMD算法存在難以選取預(yù)設(shè)尺度K和分解后的有效固有模態(tài)函數(shù)(Intrinsic Mode Function,簡稱IMF)分量的問題,提出了一種基于相關(guān)系數(shù)的VMD方法進(jìn)行參數(shù)優(yōu)化,并將其用于檢測管道泄漏信號(hào)。首先通過仿真信號(hào)驗(yàn)證基于相關(guān)系數(shù)的VMD算法的有效性;最后將基于相關(guān)系數(shù)的VMD算法應(yīng)用于管道泄漏信號(hào)檢測中,對(duì)算法的可行性進(jìn)行驗(yàn)證。3、針對(duì)VMD分解的高頻部分受噪聲干擾導(dǎo)致分解的高頻部分效果不理想的問題,提出了基于互信息的VMD方法。首先通過仿真信號(hào)驗(yàn)證基于互信息的VMD算法的有效性;最后將基于互信息的VMD算法應(yīng)用于管道泄漏信號(hào)檢測中,對(duì)算法的可行性進(jìn)行驗(yàn)證。對(duì)比VMD算法,驗(yàn)證該方法的有效性。4、針對(duì)管道不同工況信號(hào),利用基于相關(guān)系數(shù)的VMD算法選取預(yù)設(shè)尺度K值,再通過基于互信息的VMD算法進(jìn)行管道特征提取,從而進(jìn)行工況識(shí)別,通過實(shí)驗(yàn)結(jié)果驗(yàn)證該方法的準(zhǔn)確率和可行性。
[Abstract]:There are many technical means to ensure pipeline safety, among which leak detection technology is the most important and widely used. When the pipeline is running and maintaining, it is necessary to monitor the pipeline signal in real time, accurately judge the running condition of the pipeline, diagnose the hidden trouble of leakage in time, and avoid the occurrence of safety accident. The infrasonic wave signal produced by natural gas pipeline leakage belongs to non-stationary signal and is seriously disturbed by the outside world in the course of pipeline propagation, which leads to the high misjudgment rate of leakage and the occurrence of false alarm, false alarm and false alarm. Adaptive time-frequency analysis method is an important means to analyze non-stationary signals, aiming at the characteristics of infrasonic signals generated by pipeline leakage collected in laboratory. In this paper, the variable Mode decomposition (VMD) algorithm based on the correlation coefficient is used to select the preset scale K value, and then the pipeline feature is extracted by the mutual information based VMD algorithm to identify the working conditions. In order to improve the accuracy of the system to determine pipeline leakage. The main work of this paper is as follows: 1, referring to relevant papers on pipeline leakage detection, deeply understanding the relevant methods of pipeline leakage detection, selecting VMD method to be applied to natural gas pipeline leakage detection, and analyzing the fault situation of pipeline leakage detection. Classification of fault conditions, and finally determine the target type of pattern recognition; In order to solve the problem that it is difficult to select the preset scale K and the decomposed Intrinsic Mode function (IMF) component in the VMD algorithm, the signal acquisition under different pipeline operation conditions is completed. A VMD method based on correlation coefficient is proposed for parameter optimization, and it is used to detect pipeline leakage signal. At first, the validity of the VMD algorithm based on correlation coefficient is verified by simulation signal. Finally, the VMD algorithm based on correlation coefficient is applied to the detection of pipeline leakage signal. The feasibility of the algorithm is verified. A mutual information based VMD method is proposed to solve the problem that the high frequency part of VMD decomposition is not satisfactory due to the noise interference. At first, the validity of the mutual information based VMD algorithm is verified by the simulation signal. Finally, the mutual information based VMD algorithm is applied to the pipeline leakage signal detection, and the feasibility of the algorithm is verified. Comparing with the VMD algorithm, the validity of the method is verified. According to the pipeline signal under different working conditions, the default scale K value is selected by using the VMD algorithm based on the correlation coefficient, and then the pipeline feature is extracted by the VMD algorithm based on mutual information. The accuracy and feasibility of the method are verified by the experimental results.
【學(xué)位授予單位】：東北石油大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TE973.6
，

本文編號(hào)：2138356