基于BP神经网络的涪陵气田钻井周期预测方法  

Drilling Cycle Prediction Method for Fuling Gas Field Based on BP Neural Network

作　　者：黄迪箫笙 HUANG Dixiaosheng(Research Institute of Petroleum Engineering Technology,Sinopec Jianghan Oilfield Company,Wuhan,Hubei 430042,China)

机构地区：[1]中石化江汉油田分公司石油工程技术研究院,湖北武汉430042

出　　处：《江汉石油职工大学学报》2024年第1期36-38,共3页Journal of Jianghan Petroleum University of Staff and Workers

基　　金：中国石油化工股份有限公司科研项目“涪陵页岩气田提高采收率技术研究”,编号:P22183。

摘　　要：针对目前涪陵页岩气田调整开发导致的目的层埋深不同、各区块复杂情况不同等现象,利用BP神经网络算法,基于钻井现场实测数据依据互信息相关联分析优选影响钻井周期的输入参数,通过BP神经网络初始值、阈值的优化,建立出基于BP神经网络的涪陵气田钻井周期预测方法。通过施工实际数据对比,其预测误差在10%以内。To address the differences in target layer burial depth and the complexity of various blocks caused by the adjustment of development in the Fuling shale gas field,a drilling cycle prediction method based on BP neural network algorithm is proposed.It utilizes actual drilling data for optimized input parameters based on information correlation analysis.Through the optimization of initial values and thresholds of the BP neural network,a reliable drilling cycle prediction method for the Fuling gas field is established.The predicted error is within 10%based on the comparison with the field actual data.

关 键 词：钻井周期预测 BP神经网络 钻井参数 

分 类 号：TE24[石油与天然气工程—油气井工程]