岩溶塌陷预测的高斯过程机器学习模型  

Machine learning model of Gaussian process for prediction of karstic collapse

作　　者：燕柳斌[1] 张研[1] 苏国韶[1] 

机构地区：[1]广西大学土木建筑工程学院工程防灾与结构安全教育部重点实验室,广西南宁530004

出　　处：《广西大学学报（自然科学版）》2011年第1期172-176,共5页Journal of Guangxi University（Natural Science Edition）

基　　金：国家自然科学基金资助项目(51069001);广西自然科学基金(桂科自0832023)

摘　　要：为研究在多种复杂影响因素下岩溶塌陷如何准确预测,针对现有方法的局限性,建立了一种基于高斯过程机器学习的岩溶塌陷预测模型。该模型通过对少量学习样本的学习,就可以建立岩溶塌陷与其影响因素之间的复杂非线性映射关系。将模型应用于工程实例,研究结果表明,岩溶塌陷预测的高斯过程机器学习模型是科学可行的,具有预测精度高、适用性强、参数自适应化且易于实现的优点。Aiming at that it was still difficult to reasonably predict karstic collapse,a model based on machine learning of Gaussian process（GP） was proposed for prediction of karstic collapse.With few learning samples,the nonlinear mapping relationship between karstic collapse and its influence factors was established by the GP model.The model was applied to a project.The results show that the GP model is feasible,effective and simple to implement for prediction of karstic collapse.It has such merits as self-adaptive determination of parameters and excellent capability in solving non-linear small samples problems.

关 键 词：岩溶塌陷 高斯过程 机器学习 预测 

分 类 号：P642.254[天文地球—工程地质学]