| 序贯最小二乘支持向量机的结构系统识别 |
| |
| 引用本文： | 唐和生,薛松涛,陈镕,晋侃.序贯最小二乘支持向量机的结构系统识别[J].振动工程学报,2006,19(3):382-387. |
| |
| 作者姓名： | 唐和生 薛松涛 陈镕 晋侃 |
| |
| 作者单位： | 1. 同济大学结构工程与防灾研究所,上海,200092
2. 同济大学结构工程与防灾研究所,上海,200092;日本近畿大学理工学部建筑学科,日本,大阪,577-8502 |
| |
| 摘 要： | 提出一种用于结构系统识别的序贯最小二乘支持向量机（SLS-SVM）方法,通过对训练数据的序列进入和数据缩减,分别采用增量算法和减缩修剪算法有效地改进了LS-SVM.这种方法克服了标准LS-SVM算法的稀疏性缺失的缺点,并使LS-SVM的序贯训练成为可能.对非线性滞迟结构的在线参数识别显示了所提出方法的鲁棒性和高效率,同时也表明SLS-SVM算法的速度比批处理SVM算法要快.
|
| 关 键 词： | 系统识别 滞迟结构 序贯 最小二乘 支持向量机 |
| 文章编号： | 1004-4523(2006)03-0382-06 |
| 收稿时间： | 2005-06-29 |
| 修稿时间： | 2005-12-12 |
Sequential LS-SVM for structural systems identification |
| |
| TANG He-sheng,XUE Song-tao,CHEN Rong,JIN Kan.Sequential LS-SVM for structural systems identification[J].Journal of Vibration Engineering,2006,19(3):382-387. |
| |
| Authors: | TANG He-sheng XUE Song-tao CHEN Rong JIN Kan |
| |
| Affiliation: | 1. Research Institute of Structural Engineering and Disaster Reduction, Tongji University, Shanghai 200092, China; 2. Department of Architecture, School of Science and Engineering, Kinki University, Osaka 577-8802, Japan |
| |
| Abstract: | A sequential Least Squares Support Vector Machines(SLS-SVM) method is proposed for identification of structural systems in this paper.It efficiently updates a trained LS-SVM by means of incremental and decremental pruning algorithms whenever a sample is added to,or removed from,the training set.The method overcomes the drawback of sparseness lost within the standard LS-SVM and makes online training for the LS-SVM possible.Examples of nonlinear hysteretic structure parameters for online identification problems show the robustness and efficiency of the proposed method.They also show that the SLS-SVM algorithm is faster than the batch SVM algorithm. |
| |
| Keywords: | systems identification hysteretic structure sequential least squares support vector machines |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
