基于实验研究的油气钻采水平两相流管道泄漏声发射检测

Acoustic emission detection of horizontal gas-liquid two-phase flow pipeline leakage based on experimental research

气液两相流管道在油气钻采、运输中应用广泛,但对其泄漏检测的研究较少。本文通过搭建水平气液两相流泄漏实验系统,利用声发射(AE)检测原理对气体压力、流型、泄漏孔径、泄漏位置等因素对泄漏声发射信号的影响进行了实验研究,提出通过经验模态分解(EMD)去噪并用小波包分解(WPD)提取声发射信号特征输入BP神经网络进行泄漏存在性、泄漏流型识别。研究结果表明,本文提出的泄漏声发射检测方法提高了BP神经网络对气液两相流泄漏模式的识别精度并可以实现对两相流管道的有效检测,具有较好的借鉴意义。

Gas-liquid two-phase flow pipeline is widely used in oil&gas drilling,production and transportation,but there are few researches on its leakage detection.In this paper,through establishing a horizontal gas-liquid two-phase flow leakage experimental system,the influence of gas pressure,flow pattern,leak aperture,leak position and other factors on leaking acoustic emission(AE)signal was studied by using acoustic emission detection principle.Then,empirical mode decomposition(EMD)is used to denoise and wavelet packet decomposition(WPD)is adopted to extract the features of acoustic emission signal,which is input into BP neural network to identify the existence and flow pattern of leakage.The results show that the leakage detection method proposed in this paper improves the accuracy of BP neural network in detecting the leakage pattern of gas-liquid two-phase flow and realizes the effective detection of two-phase flow pipeline,which has a good reference significance.