摘要
神经网络模型是处理非线性问题较好的一种方法之一。文章通过对人工神经网络的分析,建立了瞬态振动法测定土密实度的神经网络模型,网络的学习算法采用改进的BP算法。并对建模结果的准确性和可靠性进行了验证和讨论,结果表明将神经网络应用于土密实度的定量分析问题中,效果是良好的。
ANN is one of the best method to deal with non-linear problem. By analysis of ANN, the model of BP neural network is built for soil compactness quantitative analysis. The correctness and reliability of the model are then validated and discussed. Result shows that it quite agree with the tested result by ring-knife method.
出处
《科技通报》
2007年第1期116-121,136,共7页
Bulletin of Science and Technology
