基于神经网络的虚拟基准站自差分GPS定位的研究

Study of self-difference GPS positioning by fictitious datum station based on artificial neural network

提出了一种适合于铁路和公路固定线路的虚拟基准站自差分 GPS定位的方法 ;讨论了建立动态虚拟基准站的方法及基于神经网络的虚拟基准站的结构 ,并对这一神经网络进行了训练。实验表明 ,系统的定位精度得到了提高 。

This paper presents a method of self difference GPS positioning by fictitious datum station,which is suited to settled line such as railway and highway.Some problems such as the foundation of dynamic fictitious datum station,the neural network framework of fictitious datum station are discussed.The method of neural network training is described.The experiment has testified the system positioning precision is improved,and the dynamic tracking ability is strong,too.