基于BP神经网络的伪距观测值电离层误差分离  

Pseudo-range Observations Ionospheric Error Separation Based on BP Neural Network

作　　者：李阳林[1,2] 黄文德[2] 盛利元[1] 

机构地区：[1]中南大学物理与电子学院,长沙410083 [2]国防科技大学机电工程与自动化学院,长沙410073

出　　处：《全球定位系统》2015年第6期1-5,共5页Gnss World of China

基　　金：第二代卫星导航重大专项(编号:GFZX0301010105)

摘　　要：针对单频接收机用户伪距观测值中电离层延迟误差分离难的问题,提出了基于BP神经网络的误差分离方法,以接收机连续一周的观测数据作为训练样本,将样本中时间、卫星高度角、方位角及伪距观测值作为BP神经网络的输入单元,利用CODE提供的高精度GIM计算出对应的电离层延迟误差,并将其作为输出单元对BP神经网络进行训练。最后,利用训练好的模型对接下来一周各可见星伪距观测值的电离层延迟误差进行分离。实例表明:利用BP神经网络模型对伪距观测值的电离层延迟误差进行分离,分离精度均在80%以上,最佳可达90%.该方法是从大数据的角度解决导航问题的一次有益尝试。Aiming at the problem of ionospheric delay error,proposed a new error separation method based on BP neural network.The training sample is based on a week of observation data and set the pseudo-distance ionospheric delay error separation-related factorstime,satellite elevation angle,azimuth and pseudo-range observations as BP neural network input unit,use CODE published high-precision GIM calculated corresponding ionospheric delay error as an output unit for BP neural network training.Finally,use the trained model for each satellite to separate next week ionospheric delay error of pseudorange observations.The results show that：the use of BP neural network model to separate ionospheric delay error of pseudo-range observations,both of the separation accurate is above 80%,and the best is up to 90%.The method used a large number of actual historical data,it is a good attempt to solve navigation problems from the perspective of big data.

关 键 词：BP神经网络 单频接收机 伪距 电离层延迟误差 

分 类 号：P228.4[天文地球—大地测量学与测量工程]