【摘要】 针对目前油井井下工况诊断效率低下的问题,提出了基于实时生产数据的油井工况诊断系统,采用物联网架构实现三相电参、套管压力、示功图和油井动液面等实时生产数据的采集、传输与应用,利用灰度提取方法提取示功图图像型数据的特征值,与数值型数据融合分析,增加BP神经网络的输入神经元,建立工况诊断模型。对比于通过示功图提取特征值的诊断方法,油井工况诊断系统可以大幅度提高油井工况诊断的准确性,对于实现远程智慧化油田发展具有重要意义。更多还原

【Abstract】 Aiming at the problem of low efficiency of oil well condition diagnosis, an oil well condition diagnosis system based on real-time production data is proposed. The real-time production data such as three-phase electrical parameters, casing pressure, indicator diagram and the dynamic liquid level of oil well are collected, transmitted and applied by Internet of things. The feature values of the image data of the indicator diagram are extracted by gray extraction method, and the data fusion analysis of the feature values with the numerical data of the indicator diagram is carried out, and the oil well condition diagnosis model is establish by increasing the input neurons of BP neural network. Compared with the diagnosis method of extracting eigenvalues from indicator diagram, this method can greatly improve the accuracy of oil well condition diagnosis, which is of great significance for the realization of remote intelligent oilfield.更多还原