边坡稳定性预报的MEP分类预测算法  

Forecasting Soil Slope Stability with MEP Algorithm

作　　者：姜大志[1] 彭晨枫[1] 刘亮[2] 

机构地区：[1]汕头大学工学院计算机系,广东汕头515063 [2]江苏省地质工程有限公司,江苏南京210018

出　　处：《汕头大学学报（自然科学版）》2011年第4期66-72,共7页Journal of Shantou University：Natural Science Edition

基　　金：广东高校优秀青年创新人才培养计划项目(LYM10078);汕头大学青年基金(YR10006)

摘　　要：鉴于传统方法和数值分析方法的局限性,引入了密切影响边坡稳定性的若干参数,运用多表达式程序设计方法建立边坡稳定性评价模型.基于已有的边坡事例,对该模型进行训练和自我学习,然后用样本训练后的优化模型对新的土质边坡数据集进行预测评价.数值实验表明:多表达式程序设计算法在建立边坡稳定性评价模型中具有较高的可信度.For the limitation of traditional methods and numerical analysis, the relative parameters that deeply affect the soil slope stability were introduced and the soil slope stability evaluation model was built with the MEP algorithm. The model has a great ability in dealing with nonlinear problems. The self training and learning were based on existing slope examples, after which new soil slope data could be evaluated by the optimized model from training. The numerical experiment indicated that MEP algorithm has a high reliability in building soil slope stability evaluation model.

关 键 词：边坡稳定性 MEP算法 多表达式程序设计 评价模型 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]