城市地下天然气管网作为城市“生命线”的重要组成部分，是承担能源输送任务的城市重要基础设施。随着我国经济的飞速发展和人民生活水平的不断提升，我国天然气使用普及度日益提高，城市地下天然气管线建设版图迅速扩张，然而，由于管线老化、腐蚀穿孔、第三方破坏等因素导致的天然气管道泄漏事故也频繁发生。天然气具有易燃易爆的危险特性，泄漏发生后，及时准确地定位泄漏源将为应急响应赢得宝贵时间，否则不仅会造成经济损失与环境污染，甚至可能导致重大人员伤亡事故的发生。天然气管道声波法泄漏定位准确度较高，但现有的方法多以管壁或管内介质作为声传播介质，传感器需与管道直接接触，并不适用于未安装相关传感器的现役老旧管道。为此，本文试图研究一种在地面探测孔中布置传感器，以不接触管道的方式实现泄漏源定位的方法。通过研究泄漏声源特性、泄漏声特征提取、泄漏声时延估计及泄漏源定位方法，建立一种以土壤为声传播介质的埋地天然气管道泄漏源的定位理论与方法。本文通过数值计算分析对比了不同工况下埋地天然气管道与架空天然气管道的泄漏声源特性，厘清了稳定泄漏状态下天然气管道泄漏声源的发声机理；分析了埋地管道不同工况下泄漏声源与流场之间的关系，发现流场特征参数可用来表征泄漏声源变化情况，并通过管道泄漏声源模拟实验对相关结论进行了验证。研建了埋地天然气管道泄漏模拟实验平台，在大量实验数据的基础上，分析了埋地管道的土壤内泄漏声特性，并基于其时频特性，开展了针对土壤内泄漏声信号特性的特征频段提取方法的研究，实现了对特征频段土壤内泄漏声时域信号的重构。研究了针对不同传声器间所采集土壤内泄漏声信号的时延估计方法，并利用重构后的土壤内泄漏声特征频段时域信号对所建立的时延估计方法的效果进行了验证。提出了以土壤为声传播介质的埋地天然气管道泄漏源定位方法，并通过泄漏定位实验验证了新方法的定位效果，在衰减程度较大的湿润型（含水率25%体积分数）土壤中，实验用传声器探测距离可达5m（传声器动态范围下限为19dB），定位误差处于8%-12%之间，对埋地天然气管道的泄漏定位工作具有较好的应用意义。
As an important part of the urban lifeline, the buried natural gas pipeline network is an important urban infrastructure that undertakes the task of energy transmission. With the rapid development of Chinese economy and the continuous improvement of living standards, the popularity of natural gas use in China is increasing, and the construction of urban buried natural gas pipelines is expanding rapidly. However, natural gas pipeline leakage accidents due to external interference, corrosion, construction defect, material failure and other factors also occur frequently. Since natural gas is flammable and explosive, once leak occurs, timely and accurate localizing the leak source will win valuable time for emergency response. Otherwise, it will not only cause economic losses and environmental pollution, but may even cause serious casualties. The acoustic method for leak localization of natural gas pipeline shows high localization accuracy. However, the pipe wall or the medium in the pipeline are mostly used as the sound propagation medium in existing method, and the sensors need to be in direct contact with the pipe, which is not suitable for the active and old pipelines without the relevant sensors installed. Therefore, to study the leak localization method based on the soil as the sound propagation medium for buried natural gas pipeline, this paper developed a series theoretical and experimental researches about the characteristics of leak noise at source, feature extraction of leak noise, the time delay estimation of leak noise and the leak noise localization. The main work is as follows:The characteristics of leak noise at source for buried and above-ground natural gas pipeline were analyzed and compared through numerical simulation. The generation mechanism of leak noise at source about buried natural gas pipeline in stable leak state was clarified. The relationship between the leak noise at source and the flow field under different conditions of buried pipeline was analyzed, and it is found that the parameters of the flow field can be used to characterize the change of leak noise at source. And the relevant conclusions were verified by the experiment results of leak noise at source about buried natural gas pipeline.The experimental platform of leak simulation for buried natural gas pipeline was established. And the characteristics of leak noise in the soil collected by the experiment under different leak conditions were analyzed. Based on its time-frequency characteristics under different conditions, the study of the characteristic frequency band extraction method about leak noise in soil were conducted to realize the reconstruction for the time-domain signal about the characteristic frequency band. In addition, related research has been carried out on the time delay estimation method for the leak noise in soil collected by the different microphones. And the effect of the time delay estimation method is verified by the reconstructed time-domain signal of the leak noise in soil at the characteristic frequency band.The leak localization method based on the soil as the sound propagation medium for buried natural gas pipeline is proposed. And the leak localization effect of the novel method is verified by leak localization experiments, it shows the localization error under relevant experimental conditions is between 8%-12%. Moreover, the application of the novel leak localization method is evaluated based on the sound propagation characteristics of the soil, and the maximum detection distance can reach 5 m (the lower limit of the microphone dynamic range is 19dB) in the moist soil with the greatest attenuation (25% volume fraction). It has good application significance for the leak localization of buried natural gas pipelines.
