School of Petroleum Engineering,China University of Petroleum (Beijing)MOE Engineering Research Center ofGas Energy Development and Utilization,China University of Petroleum(Beijing)Oil and Gas Engineering Division,Liaohe Petroleum Exploration Bureau

针对目前采用单一数据源进行煤层气水平井随钻测量与判识的局限和不足,提出随钻测量与判识技术是煤层气水平井钻井施工过程中实现地质导向的关键,是钻头在目的层中顺利延伸的有力保障,充分利用多源信息融合技术处理不同信息源的优势,以BP神经网络作为融合算法,将综合录井数据与元素随钻测量技术进行融合,优选出与多源信息融合技术配套的元素随钻测量仪器,并设计了随钻测量仪器联机方案,确定了随钻判识功能实现的技术流程,建立了多源信息融合随钻判识网络。在沁水盆地某井利用基于多源信息融合的随钻判识网络进行实际应用可知,其可得到准确的判识结果。

According to the lhit and shorage of a single data source applied to the measurement and identifcation of thecoalbed methane horizontal well while drilng the paper provided that he measurement and dentification tfchnology was the key to ealize the geological seeing during the drlling construction process of thecoalbed methane horizontal ell and was the power suport of the drlin bt to be succssilly expanded in the target stata.with a ful iliration on the advantages ofthe muiti-nformation integration technology to pro.ess difrent information sources.based on theBP neural network as the inte ration agrfthm,the comprehensive lo:ging data and the elements measuring technology while drling were inegrated.The element measuring instrument whileadrling matc hed to the mut-information integation technology was optimized.An onine plan of the measuring instruction whiedriling was designed a technical prcedure torealize the identificatin funcion whie&rilling was determined and an identification networs of the maulit-information integration was establis ed.An acual appication of the identfration netwvork whie diling o a well in Qinshui Basin based on the multi-information integration showed that the accurate identfication results were accurately obtained.