地层真电阻率恢复神经网络法

水驱油田进入高含水阶段后,利用测井信息恢复地层真电经是一个关键问题。文中首先利用自回归线性预测技术消除地层厚度对测井曲线值的影响,然后从多条测井曲线提取有效特征,利用取心井训练神经网络,进而预测未知井,达到地层真电经恢复的目的。实验资料的处理结果表明,此法具有好的应用前景。

Restoration of True Formation Resistivity by Neural Network

When water driving oilfield enters into high watercut stage, it becomes a key issue to use well logging information to restore true formation resistivity. Auto regressive linear prediction technique is used to eliminate the influence of stratum thickness on well logging data. And then effective characters are extracted from several logging curves and core data are used to train neural network to predict the formation resistivity of unknown well, from which true resistivity is restored. The processing results of actual well logging data show that this technique has good prospects.