摘要
提出应用自组织特征映射(SOFM)神经网络进行砂土液化评价,根据实测资料和行业规范,建立具有7个输入参数,4个输出类别的SOFM神经网络模型对砂土液化的严重程度做出评价.实例研究表明,应用SOFM神经网络评价砂土液化高效可行,为砂土液化的研究提供了新方法.
<Abstrcat> Self-organization Feature Mapping(SOFM) neural network is applied to evaluate sand liquefaction potential. Based on the case-historic data and the pr ofessional norms, the SOFM network model with seven input parameters and four ou tput sorts is set up to evaluate sand liquefaction potential. The testing result s of practical examples show that the method based on SOFM neural network for ev aluating sands liquefaction potential is feasible and effective and has gain fin e results, which provides a new approach to research sands liquefaction poten tial.
出处
《华中科技大学学报（城市科学版）》
CAS
2005年第2期23-26,共4页
Journal of Huazhong University of Science and Technology
基金
国家自然科学基金资助项目(69974026).
关键词
自组织特征映射(SOFM)
神经网络
砂土液化
评价
self-organization feature mapping(SOFM)
neural network
sands liquefactio n potential
evaluation
