基于卷积神经网络的在线泵效计算方法研究与应用  

Research and Application of Online Pump Efficiency Calculation Method Based on Convolutional Neural Network

作　　者：陈佳乐 高鹏 朱丹丹 金学锋 朱丽萍[1] 张战敏 Chen Jiale;Gao Peng;Zhu Dandan;Jin Xuefeng;Zhu Liping;Zhang Zhanmin(School of Information Science and Engineering,China University of Petroleum(Beijing),Beijing,102249,China;PetroChina Huabei Oilfield Company Engineering and Technology research Institute,Renqiu,062550,China)

机构地区：[1]中国石油大学(北京)信息科学与工程学院,北京102249 [2]中国石油华北油田公司工程技术研究院,河北任丘062552

出　　处：《石油化工自动化》2021年第1期8-11,共4页Automation in Petro-chemical Industry

摘　　要：油井泵效的计算是油田生产管理中的一项重要工作,针对传统泵效计算方法计算复杂,速度慢的问题,提出一种基于卷积神经网络的泵效计算方法。该方法通过卷积操作从地面示功图中提取特征图,并融合油井参数和抽油机参数,挖掘多源参数与泵效之间的非线性关系。此外,将井下泵示功图作为先验知识引入,有助于泵效计算精度的提升。应用测试结果表明,基于卷积神经网络的泵效计算方法精度高、速度快,可支持多井在线计算,符合油田数字化的需求。The calculation of oil well pump efficiency is an important work in oilfield production management. Aiming at the problem of complex and slow calculation by using the traditional method, a pump efficiency calculation method based on the convolutional neural network is proposed. The method extracts the feature chart from the ground indicator diagram by the convolution operation, with integrating parameters of the well and pumping unit to excavate the nonlinear relationship between the multi-source parameters and pump efficiency. In addition, the introduction of downhole pump indicator diagram as the prior-in-use knowledge is helpful to improve the accuracy of pump efficiency calculation. The practical application results show the pump efficiency calculation method based on the convolution neural network has high accuracy and fast speed, which can support multi-well online calculation, and satisfy the needs of oilfield digitization.

关 键 词：卷积神经网络 泵效 地面示功图 抽油机井 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]