概率神经网络方法在岩性识别中的应用

The Probability Neural Networks for Lithology Identification

本文研究利用概率神经网络方法进行测井资料的岩性识别;建立了测井解释的岩性识别模型,并利用该模型对测试样本进行预测,预测结果与实际测量结果相比具有较好的一致性,其计算量小且预测精度与收敛速度较BP神经网络模型有了很大的提高;应用表明,概率神经网络在岩性识别问题中有着一定的应用前景。

Lithologic identification from well-logging information based on PNN （Probability Neural Network） is studied in this paper. Lithologic identification model for well-logging interpretation is built and applied to predict the testing samples. The prediction result has higher consistency with the practical cases. The prediction precision and convergence rate is greatly improved compared to the traditional BP Neural Networks, and the computational complexity is also greatly reduced. The results obtained show that the PNN is very promising in lithologic identification.