摘要
高程异常值的快速、准确获取是GPS高程数据得到应用的前提。本文提出一种基于RBF神经网络的GPS高程拟合方法,针对RBF基函数中心的确定以及网络模型参数的选择问题,采用K-means算法自适应的实现对RBF基函数中心的选择,在此基础上利用AIC准则完成对网络模型参数的自动寻优,最后采用实际数据对所提方法进行验证,并与传统BP神经网络拟合方法进行对比,实验结果表明所提方法可以获得更好的拟合精度。
The fast and accurate acquisition of elevation abnormal values is the prerequisite for the application of GPS elevation data.This paper proposes a GPS elevation fitting method based on RBF neural network.Aiming at the determination of the RBF basis function center and the selection of network model parameters,the K-means algorithm is used to adaptively realize the selection of the RBF basis function center.On this basis,it uses the AIC criterion to complete the automatic optimization of the network model parameters.Finally,the actual data is used to verify the proposed method,and compared with the traditional BP neural network fitting method.The experimental results show that the proposed method can achieve a better fitting accuracy.
作者
丁凯
王泽林
DING Kai;WANG Zelin(Guangdong Underground Pipe Network Engineering Survey Company,Guangzhou 510000,China)
出处
《测绘与空间地理信息》
2023年第5期100-103,107,共5页
Geomatics & Spatial Information Technology
