基于AFSA-LSSVM的位移反分析方法研究  

Research on Back Analysis Method of Displacement Based on AFSA-LSSVM

作　　者：张凌然[1,2] 徐卫亚[1,2] 聂卫平[1,2] 石崇[1,2] 杨云浩[1,2] 

机构地区：[1]河海大学岩土力学与堤坝工程教育部重点实验室,江苏南京210098 [2]河海大学岩土工程科学研究所,江苏南京210098

出　　处：《水电能源科学》2011年第4期78-80,共3页Water Resources and Power

基　　金：国家自然科学基金资助项目(50911130366;50979030);国家科技支撑计划基金资助项目(2008BAB29B01)

摘　　要：以糯扎渡水电站工程为例,依据LSSVM原理构建了待反演参数与位移的预测模型,采用AFSA优化LSSVM模型参数提出基于AFSA-LSSVM的位移反分析方法,寻求预测位移值与监测位移值最接近的适应值,以获得最真实的待反演参数。结果表明,该方法较PSO-LSSVM能搜索到更优适应值,计算精度高,适用性较强。Taking Nuozhadu hydropower station project for an example,this paper proposes a forecasting model for inversion parameters and displacement based on LSSVM theory.And the AFSA method is applied to optimize parameters of the LSSVM model.Then back analysis of displacement is adopted with AFSA-LSSVM method.At the same time,the exact parametric inversion is obtained by seeking the fitness value which makes the forecasting displacement and monitoring displacement to be closed.The results show that the proposed method can obtain more optimal value than that of PSO-LSSVM,and it has high accuracy and strong adaptability.

关 键 词：位移反分析 最小二乘支持向量机 鱼群算法 糯扎渡水电站 

分 类 号：TV698.11[水利工程—水利水电工程] TU45[建筑科学—岩土工程]