径向基函数神经网络在孔隙度预测中的应用

利用径向基函数神经网络对新疆乌尔禾地区三叠系白口泉组储层孔隙度进行预测,取得了较好的效果。对比分析结果表明,径向基函数神经网络预测结果较BP神经网络以及线性回归法预测结果精度高。阐述了应用中应该注意的问题。

An application of RBF neural network in reservoir porosity prediction

The Radial Basic Function(RBF) neural network has been used to predict reservoir porosity of Triassic Baikouquan Formation in Wuerhe area,Xijiang Region,leading to quite good results.By a comparable analysis,it is clear that RBF neural network is higher in prediction accuracy of reservoir porosity than BP neural network and the linear regression method.In addition,several issues that should be emphasized in application of the technique are expressed.