基于RE-CFSFDP与DSA-LSSVM的山脊梁数据预测方法研究  

The Research of Ridge Data Prediction Method Based on RE-CFSFDP and DSA-LSSVM

作　　者：杨本钊 YANG Benzhao(Master of Commerce,The University of Sydney,Sydney)

机构地区：[1]悉尼大学商学院,悉尼

出　　处：《长春理工大学学报（自然科学版）》2022年第2期130-137,共8页Journal of Changchun University of Science and Technology(Natural Science Edition)

摘　　要：面向真实山体数据(山脊梁变形位移数据为主,边坡地下水位等20项相关数据为辅)进行分析,在利用DSA优化最小二乘支持向量机(LSSVM)的核心参数的方法基础上,提出响应效率(Response Efficiency)概念,利用CFSFDP聚类方法以不确定数据降雨量为核心数据,将其他数据分为少雨、中雨、多雨三个不同时间分组数据,再对不同时间区域的数据进行预测,使有了时间延迟的聚类数据在DSA-LSSVM模型上有明显改善。In this paper,the real mountain data(mainly the deformation and displacement data of mountain ridge,supplemented by 20 related data such as slope groundwater level)are analysed. based on the method of optimizing the core parameters of least squares support vector machine(LSSVM) by DSA,the concept of Response Efficiency is put forward,and the uncertain data rainfall is taken as the core data by CFSFDP clustering method. Other data are divided into three different time groups:less rain,moderate rain and more rain,and then the data in different time regions are predicted,which makes the clustering data with time delay significantly improved on DSA-LSSVM model.

关 键 词：DSA-LSSVM 不确定数据 聚类 RE-CFSFDP 

分 类 号：N32[自然科学总论]