基于门控循环单元网络的钻井井漏智能监测方法  

Intelligent monitoring method for drilling lost circulation based on GRU network

作　　者：李辉[1] 刘凯[2] 李威桦 孙伟峰[3] 戴永寿[3] LI Hui;LIU Kai;LI Weihua;SUN Weifeng;DAI Yongshou(Well-Tech R&D Institutes,China Oilfield Services Limited,Langfang 065201,China;College of Control Science and Engineering,China University of Petroleum(East China),Qingdao 266580,China;College of Oceanography and Space Informatics,China University of Petroleum(East China),Qingdao 266580,China)

机构地区：[1]中海油田服务股份有限公司油田技术研究院,河北廊坊065201 [2]中国石油大学(华东)控制科学与工程学院,山东青岛266580 [3]中国石油大学(华东)海洋与空间信息学院,山东青岛266580

出　　处：《电子设计工程》2024年第3期31-36,共6页Electronic Design Engineering

基　　金：国家自然科学基金资助(62071493)。

摘　　要：井漏是钻井过程中常见的钻井风险,若对该风险发现、处理不及时,极易导致井塌事故,轻则延长施工周期,重则危害现场人员人身安全。为了提高油气井钻井过程中井漏风险识别的准确性,降低风险识别对人为经验的依赖,结合钻井参数的非线性以及长时依赖特征,提出了一种基于门控循环单元(Gated Recurrent Unit,GRU)网络的井漏风险智能识别方法。该模型以池体积、出口流量和立管压力作为监测参数构建GRU网络,能够提取监测参数的时间序列特征,以实现对井漏风险的准确识别。利用现场实测钻井数据对模型进行了实验测试,结果表明,该方法对井漏风险的识别准确率达到了90.1%,优于长短期记忆网络的识别结果。Lost circulation is a common risk during drilling process.If the risk is not discovered and dealt with in time,it is easy to lead to well collapse accident,which may prolong the construction period,or even endanger personal safety.The lost circulation monitoring methods including artificial inspection and expert system used in drilling field often have some problems,such as not finding risks in time and low identification accuracy,etc.In order to improve the accuracy of loss circulation monitoring,considering the nonlinearity of drilling parameters and the characteristics of long⁃term dependence,an intelligent lost circulation monitoring model based on Gated Recurrent Unit(GRU)neural network is proposed.Pit volume,outlet flow and standpipe pressure were used as monitoring parameters to build a GRU network model,which was used to extract the time sequence characteristics of monitoring parameters to achieve accurate identification of loss risk.The model was tested experimentally using actual drilling data from the field,the experimental results show that the accuracy of the proposed model is 90.1%,which is better than that of the Long Short⁃Term Memory network.

关 键 词：钻井安全 井漏监测 时序特征 门控循环单元网络 

分 类 号：TN98[电子电信—信息与通信工程]