黄河堤防边坡稳定性多因素敏感性分析  

Analysis on Multi-Factor Sensibility of Slope Stability of the Yellow River Embankments

作　　者：杨继红[1] 刘汉东[2] 秦四清[1] 路新景[3] 王媛媛[1] 孙强[1] 

机构地区：[1]中国科学院地质与地球物理研究所,北京100029 [2]华北水利水电学院,河南郑州450011 [3]黄河勘测规划设计有限公司,河南郑州450003

出　　处：《人民黄河》2008年第1期15-17,21,共4页Yellow River

摘　　要：利用人工神经网络Levenberg-M arquardt优化算法,结合正交试验设计理论,建立了堤防边坡稳定性的多因素敏感性分析模型,并以黄河开封段大堤为例进行了敏感性分析,结果表明:①把神经网络和正交试验结合起来是边坡稳定性多因素敏感性分析的一种有效途径;②影响边坡稳定性因素的敏感性大小依次为洪水位、黏聚力、地震加速度、内摩擦角、堤身隐患和容重。It builds up multi-factor sensibility analysis model for slope stability of embankments by using Levenberg-Marquardt optimized calculation method of artificial neural network and combining with the design theory of orthogonal test and conducts sensibility analysis by taking Kaifeng section of the Yellow River as an example.The outcomes show that a) combining neural network and orthogonal test is an effective way for multi-factor sensibility analysis of slope stability and;b) the effected sensibility degree of the slope stability factor is in proper order of flood water level,adhesion,earthquake acceleration,angle of internal friction,hidden defects of embankments and volumetric weight.

关 键 词：正交试验 非线性映射 人工神经网络 敏感性分析 堤身隐患 堤防 黄河 

分 类 号：TV871[水利工程—水利水电工程] TV882.1