油井工况智能监测与风险防控技术研究  

Intelligent Monitoring and Risk Prevention and Control Technology of Oil Well Condition

作　　者：王勇[1] WANG Yong(National Academy of Economic Security,Beijing Jiaotong University,Beijing 100044,China)

机构地区：[1]北京交通大学国家经济安全研究院,北京100044

出　　处：《河南科学》2020年第1期63-68,共6页Henan Science

基　　金：中国石化科技攻关项目(LMKJ201719)。

摘　　要：为了实现油井工况的自动评估与智能管理,达到防范风险的目标,基于深度学习技术建立了油井工况智能监测与风险防控方法.将实际油田数万条示功图数据整理为卷积神经网络模型的训练集,并基于该训练集,利用监督误差反向传播算法对基于卷积神经网络模型的示功图识别模型进行了训练和测试.结果表明,基于深度学习技术的示功图识别模型的准确度达到95%以上.研究成果应用于国内某油田,采油系统效率总体提升2.67%,工况合格率提升11%,防范风险成功率提高60%.研究成果为同类油田提供了一定的技术借鉴.In order to realize the automatic evaluation and intelligent management of oil well conditions and achieve the goal of preventing risks,based on deep learning technology,the intelligent monitoring and risk prevention and control methods for oil well conditions are established.The tens of thousands of dynamometer data of the actual oilfield are organized into the training set of the convolutional neural network model.Based on the training set,the dynamometer recognition model based on the convolutional neural network model is trained and tested by the supervised error back propagation algorithm.The results show that the accuracy of the dynamometer recognition model based on deep learning technology is over 95%.The research results are applied to a domestic oilfield.The efficiency of the oil production system is generally increased by 2.67%,the qualified rate of working conditions is increased by 11%,and the success rate of risk prevention is increased by 60%.The research results provide a certain technical reference for similar oil fields.

关 键 词：风险防范 生产工况 深度学习 卷积神经网络 

分 类 号：TE358[石油与天然气工程—油气田开发工程]