滑坡残余强度预测  

Prediction of landslide residual strength

作　　者：汤罗圣[1] 殷坤龙[1] 刘艺梁[1] 吴益平[1] 

机构地区：[1]中国地质大学(武汉)工程学院,湖北武汉430074

出　　处：《中南大学学报（自然科学版）》2013年第3期1116-1121,共6页Journal of Central South University:Science and Technology

基　　金：国家自然科学基金资助项目(40872176;41002103);中央高校基本科研业务费专项资金优秀青年教师基金资助项目(CUGL100213);教育部长江三峡库区地质灾害研究中心开放基金资助项目(TGRC2010121)

摘　　要：以滑坡基本物理参数为基础,从区域统计规律出发,采用SPSS数学分析软件对万州区滑坡残余强度的相关影响因素进行筛选,然后以重庆市万州区典型滑坡——万泉滑坡为例,分别采用神经网络模型和多元回归模型对该滑坡的残余强度进行预测,最后对其预测结果进行对比分析。研究结果表明:多元回归模型的预测精度为10%～40%,而神经网络模型的预测精度基本在5%左右,故神经网络模型预测精度更高。Based on landslide physical and mechanical parameters,the mathematical analysis software SPSS was adopted to select the related influence factors for the landslide residual strength in Wanzhou district from regional statistical rule. The neural network model and multiple regression model were applied to predict the residual strength of the typical landslide example i.e.Wanquan landslide in Wanzhou district,Chongqing.Finally,the forecast results were analyzed.The results show that the prediction accuracy of multiple regression model and neural network model are 10%-40%and 5%, respectively,and so the neural network model has a higher forecast precision.

关 键 词：滑坡残余强度 神经网络模型 多元回归模型 万泉滑坡 重庆市万州区 

分 类 号：TU398[建筑科学—结构工程]