摘要
分析井筒流量、密度、黏度等主要因素对连续流量计启动速度影响的基础上,将BP神经网络应用于连续流量计启动速度的预测。在BP神经网络预测模型中,将井筒实测流量、流体黏度和混合流体密度作为输入参数,根据不同流量、流体黏度和流体密度条件下的连续流量计启动速度等参数,建立连续流量计启动速度的BP神经网络预测模型。该BP神经网络预测模型训练后总拟合优度0.917,所得预测结果的最大相对误差为2.0%。结果表明模型的预测值与实测值吻合较好,对于连续流量计启动速度的预测具有较高的准确性和较强的实用性。
Based on the analysis of the influence of main factorson the starting speed of continuous flowmeter, such as flow rate, fluid density and viscosity, BP neural network is applied to the prediction of the starting speed of continuous flowmeter. In the BP neural network prediction model, the measured wellbore flow rate, fluid viscosity and mixed fluid density are taken as input parameters.Based on the continuous flowmeter starting speeds under different flow rate, fluid viscosity and fluid density, BP neural network prediction modelis established. After training,the BP neural network prediction model has a total goodness of fit of 0.917, and has the maximum relative error of the predicted results of 2.0%. The study results show that the predicted values of the model are in good consistency with the measured values, presenting higher accuracy and strong practicability of the prediction of the continuous flowmeter starting speed.
作者
陈科贵
毛晨飞
董军
陈愿愿
黄长兵
王小准
CHEN Kegui;MAO Chenfei;DONG Jun;CHEN Yuanyuan;HUANG Changbing;WANG Xiaozhun(School of Geoscience and Technology,Southwest Petroleum University,Chengdu,Sichuan 610500,China;Huabei Branch,China Petroleum Logging CO.LTD.,Renqin,Hebei 062550,China;Geophysical Exploration Company,Chuanqing Drilling Engineering Company Limited,Chengdu,Sichuan 610213,China;Research Institute of Petroleum Exploration and Development,Zhongyuan Oilfield,Zhengzhou,Henan 450000,China;Changqing Division of China Petroleum Logging CO.LTD.,Xi＇an,Shaanxi 710100,China)
出处
《测井技术》
CAS
CSCD
2018年第4期444-448,共5页
Well Logging Technology
基金
国家自然科学基金项目四川盆地油钾兼探的地球物理评价方法研究(编号:41372103)资助
关键词
生产测井
BP神经网络
连续流量计
启动速度
井筒
production logging
BP neural network
continuous flowmeter
starting speed
wellbore
