基于最小二乘支持向量机和人工鱼群算法的GPS高程异常拟合技术  

GPS Elevation Abnormal Fitting Technique based on Least Squares Support Vector Machine and Artificial Fish Swarm Algorithm

作　　者：宋亚宏 岳建平[1] 李静[1] 刘斌[1] 

机构地区：[1]河海大学地球科学与工程学院,南京市210098

出　　处：《勘察科学技术》2016年第4期19-21,共3页Site Investigation Science and Technology

摘　　要：为了有效地解决最小二乘支持向量机(LSSVM)进行高程异常拟合时,人工选择参数造成拟合精度损失的问题。该文采用人工鱼群算法(AFSA)对参数进行优化选择,提高了最小二乘支持向量机拟合精度。通过分析,AFSA-LSSVM模型在样本点较少时拟合残差均小于5mm,差值标准差为2.6mm,表明该算法优于传统的拟合方法。In order to effectively solve the problem of fitting accuracy loss caused by artificial selection parameters during using the least squares support vector machine（ LSSVM） to carry out the height anomaly fitting,this paper uses the artificial fish swarm algorithm（ AFSA） to optimize and select the parameters,and it improve the fitting precision of least squares support vector machine. Through analysis,the fitting residuals of the AFSA-LSSVM model are all less than 5mm when the sample points are less,and the standard deviation of difference is 2. 6mm. These show that the algorithm outperforms the traditional fitting method.

关 键 词：人工鱼群算法 最小二乘支持向量机 高程拟合 高程异常 

分 类 号：P228.4[天文地球—大地测量学与测量工程] TP18[天文地球—测绘科学与技术]