The application of improved BP neural network algorithm in lithology recognition

Abstract

Traditional technology of lithology identification bases on statistical theory, such as regression method and cluster method, which has some shortcomings. The standard BP neural network algorithm has some disadvantages like slow convergence speed, local minimum value which results in the loss of global optimal solution. BP neural network algorithm on the basis of improved variable rate of momentum factor can effectively overcome these disadvantages. Practical application shows that this method has the feature as high recognition precision and fast recognition rate so that it is suitable for recognition of lithology, lithofacies and sedimentary facies as well as geological research like deposit prediction and rock and mineral recognition. © 2008 Springer Berlin Heidelberg.