摘要
在深水、深井和超深井油气勘探领域,油气井固井施工面临着作业危险高、劳动强度大等多重挑战,导致油气井固井施工参数监测与进度预测难;为了解决这些问题,对基于云边协同和深度学习的油气井固井施工关键参数监测与进度预测进行了研究;通过云边协同组网,在现场采集和存储固井流量、压力、温度等数据,并利用MQTT轻量化通讯协议网络进行远程传输;研究基于CNN-BiLSTM-Attention网络的油气井固井施工进度预测数学模型,通过CNN网络提取油气井固井施工进度的关键特征要素,基于BiLSTM挖掘关键特征要素之间的关联关系,运用Attention机制对重要特征进行权重分配,以便抽取出更加关键及重要的油气井固井施工进度信息;经实验测试实现了油气井参数监测与预测的功能,表明所提方法具有明显的预测精度优势,且云边协同平台可以实时反映油气井固井施工过程中的各项关键参数。
In the fields of deep water,deep well,and ultra deep well oil and gas exploration,oil and gas well cementing construction faces multiple challenges such as high operational risks and high labor intensity,which makes it difficult to monitor and predict the parameters and progress of oil and gas well cementing construction;To address these issues,this paper studies a key parameter monitoring and progress prediction for oil and gas well cementing construction based on cloud edge collaboration and deep learning;Through cloud edge collaborative networking,data such as cementing flow rate,pressure,and temperature are collected and stored on-site,and remote transmission is carried out using message queuing telemetry transport(MQTT)lightweight communication protocol network;A mathematical model is presented to predict oil and gas well cementing construction progress based on convolutional neural network bi-directional long short-term memory attention(CNN-BiLSTM-Attention)network,the key feature elements of oil and gas well cementing construction progress are extracted through convolutional neural network(CNN)network,mining the correlations between key feature elements based on bi-directional long short-term memory(BiLSTM),and an Attention mechanism is used to allocate weights to important features and extract more critical and important information about oil and gas well cementing construction progress;Through experimental testing,the proposed method achieves the functions of monitoring and predicting oil and gas well parameters,with significant advantages in prediction accuracy,and the cloud edge collaborative platform can display various key parameters in real-time during the cementing process of oil and gas wells.
作者
田军政
谢雄武
钱坤
刘长春
马业
TIAN Junzheng;XIE Xiongwu;QIAN Kun;LIU Changchun;MA Ye(China Oilfield Services Limited,Langfang 065000,China;College of Mechanical and Electrical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)
出处
《计算机测量与控制》
2025年第2期54-62,共9页
Computer Measurement &Control
基金
中国航天科工集团基础科研项目(SCA24003)。
