泥化夹层残余强度的支持向量机预测  

Residual strength prediction of siltized intercalations by support vector machine

作　　者：陆有忠[1] 高永涛[1] 吴顺川[1] 王锦山[2] 崔波[1] 

机构地区：[1]北京科技大学金属矿山高效开采与安全教育部重点实验室 [2]河北科技师范学院学报编辑部,河北秦皇岛066004

出　　处：《辽宁工程技术大学学报（自然科学版）》2008年第1期45-47,共3页Journal of Liaoning Technical University (Natural Science)

基　　金：教育部重点资助项目(105016);科技部国家科技攻关计划资助项目(2004ASA615A-05)

摘　　要：基于夹于互层内的泥化夹层残余强度关系的复杂性和非线性等特点,利用支持向量机的结构简单,学习性能出色和有较强的推广性等优点,采用支持向量机方法进行预测,通过对已有实例样本的学习,建立了泥化夹层残余强度与其各影响因素之间的高度非线性映射关系,然后用样本学到的新关系预测新的泥化夹层残余强度。结果表明:与传统的方法相比,支持向量机方法精度较高,相对误差小,为预测泥化夹层残余强度提供了一条新途径.Based on the complexity and nonlinearity＇s characteristic of the residual strength of siltized intercalations which are held between interbedded, we predict the residual strength by SVM method, which is our newly-learnt knowledge. With the study of sample,the non-linear mapping relation can be established between the residual strength of siltized intercalations and all affecting factors,and then the residual strength of siltized intercalationscan can be predicted by the newly-learnt knowledge. The results show that the SVM is more precise in measuring the residual strength of siltized intercalations than traditional methods and the SVM method has advantages in its simple structure, excellent capability instudying and good application prospects,also it provids us with a novel method of measuring the residual strength of siltzed intercalations.

关 键 词：岩土力学 泥化夹层 残余强度 支持向量机 

分 类 号：TD325[矿业工程—矿井建设]