摘要
针对地质勘查中如何确定土的力学参数问题,提出了一种基于神经网络的新方法。考虑到土的物理参数测定方法比较简单,且实测变异性小,而力学参数实测变异性大的特点,利用神经网络有较好的数值逼近能力的特点,建立了回归型RBF神经网络模型来逼近两者之间的函数关系。通过实例分析,该方法可以有效地反演力学参数。
A new method based on ANN is put forward in this paper aiming to solve the problem of determining mechanics parameters of soil in geologic examination.It is easy to determine the physics parameters of clay and the measure result's variability is small,whereas,the variability of mechanics parameters is large.So a nonlinear relationship between mechanics parameters and physics parameters of clay is established.A regressive RBF neural network model is established to approach the function relationship of the two kinds of parameters.The example proves that this method is effective to reflect the mechanics parameters according to physics parameters.
出处
《黄石理工学院学报》
2007年第2期36-39,共4页
Journal of Huangshi Institute of Technology
关键词
力学参数
数值逼近
回归型RBF神经网络
反演
mechanics parameters
numerical approximation
regressive RBF neural network
reflection