摘要
为了提高GPS高程测量的精度,提出了基于BP神经网络的GPS高程拟合方法,并以2座特大桥控制网数据为例,与常规多项式曲面拟合方法进行了比较．理论和实例证明,利用BP神经网络进行GPS高程拟合是可行的,尤其是在已知点较少的情况下,该方法具有实际意义．
In order to provide an effective method of GPS height fitting, a novel method based on BP artificial neural network was proposed, and compared with the polynomial surface fitting method by taking the data of two engineering control networks as examples. TheoRy and practice show that GPS height fitting with the method of BP artificial neural network is feasible, and especially in cases with less known points, it has great practical significance.
出处
《西南交通大学学报》
EI
CSCD
北大核心
2007年第2期148-152,共5页
Journal of Southwest Jiaotong University
关键词
BP神经网络
GPS高程
高程异常
拟合
BP artificial neural network
GPS height
height abnormity
fitting
