| 基于差异进化支持向量机的坑外土体沉降预测 |
| |
| 引用本文： | 崔铁军,马云东.基于差异进化支持向量机的坑外土体沉降预测[J].中国安全科学学报,2013(1):83-89. |
| |
| 作者姓名： | 崔铁军 马云东 |
| |
| 作者单位： | 辽宁工程技术大学安全科学与工程学院;大连交通大学辽宁省隧道与地下结构工程技术研究中心 |
| |
| 基金项目： | 国家自然科学基金资助(51050003);辽宁省自然科学基金资助(201202022);大连市科技计划项目(2011E15SF118) |
| |
| 摘 要： | 就用支持向量机(SVM)预测基坑外土体沉降而言,通过差异进化(DE)算法构造适合的决策函数十分重要。在确定坑外土体沉降函数的基本形式下,进行参数反演。后将得到的解析式作为SVM的决策函数,再进行核函数转换,从而使SVM的曲线拟合更加快速,预测更加准确。对大连地铁湾家车站基坑坑外土体的沉降数据的分析及预测的结果表明,使用SVM-DE算法在计算数据量、计算消耗时间和预测精度方面优于2种方法单独使用。
|
| 关 键 词： | 坑外土体沉降 支持向量机(SVM) 差异进化(DE)算法 拟合决策函数 沉降预测 |
Prediction of Soil Settlement Outside Pit Based on DE and SVM |
| |
| CUI Tei-jun,MA Yun-dong.Prediction of Soil Settlement Outside Pit Based on DE and SVM[J].China Safety Science Journal,2013(1):83-89. |
| |
| Authors: | CUI Tei-jun MA Yun-dong |
| |
| Institution: | 1 College of Safety Science and Engineering,Liaoning Technical University,Fuxin Liaoning 123000,China 2,Center of Liaoning Tunnel & Underground Structure Engineering,Dalian Jiaotong University, Dalian Liaoning 116028,China) |
| |
| Abstract: | It is held by the authors that in order to predict the soil settlement outside pit with SVM,finding its proper decision function by means of DE algorithm is important.This paper uses DE algorithm,with understanding the basic form of soil settlement function outside pit,to make the parameter inversion,then take analytic expression obtained as SVM decision function.Finally,SVM carries on the kernel function transformation,making the SVM curve fitting more rapid and more accurate.Based on the data of soil settlement outside pit,Wanjia station in No.2 line of Dalian Metro,the analysis and prediction prove the method correctness,the results show that SVM-DE algorithm is better than the two methods used alone,in the amount of data,time consuming and prediction accuracy. |
| |
| Keywords: | soil settlement outside pit support vector machine(SVM) difference evolutionary(DE) algorithm fitting decision function settlement prediction |
| 本文献已被 CNKI 等数据库收录！ |
|
