关键词: 岩体质量分级, 神经网络, 模式分类, 比较

Abstract: On the basis of using the same training and testing data, through simulating calculation done in MATLAB artificial neural network toolbox, similarities and differences, advantages and disadvantages of BP neural network, probabilistic neural network, learning vector quantization neural network and Elman neural network on pattern classification aspect were compared. Applicable conditions of these four kinds of neural network were analyzed. It provides useful reference for artificial neural network method in the application of rock mass quality classification.

Key words: Rock mass quality classification, Neural network, Pattern classification, Comparison