Abstract: ANN（Artificial Neural Network） can make the best approximation of the actual problems through a large number of learning and training.Based on the modeling data of coal seam，the structure of ANN is designed to interpolate the interface of coal seam.Using Matlab Neural Network toolbox，the model of coal seam interface interpolation of ANN is established.The ANN is trained by a mine borehole data，then the coal seam interface interpolation is realized by the ANN.The interpolation result of the coal seam interface is 3D displayed by Matlab.

摘要： 人工神经网络方法通过大量的样本学习和训练，能够对实际问题进行最佳逼近。在分析煤层界面建模数据的基础上，设计插值煤层界面的人工神经网络结构。利用Matlab的神经网络工具箱建立对煤层界面插值的人工神经网络模型。以某矿的钻孔数据为样本对人工神经网络进行训练，实现了对煤层界面网格的插值；通过Matlab实现了对煤层界面的插值结果的三维显示。