| 卫星钟差单差的小波神网络预报 |
| |
| 引用本文： | 熊红伟,程新文,张海涛,马海荣.卫星钟差单差的小波神网络预报[J].测绘科学,2017,42(9). |
| |
| 作者姓名： | 熊红伟 程新文 张海涛 马海荣 |
| |
| 作者单位： | 1. 中国地质大学(武汉),武汉430074;中冶集团武汉勘察研究院有限公司,武汉430080;2. 中国地质大学(武汉),武汉,430074 |
| |
| 摘 要： | 针对现有卫星钟差预报模型对非平稳过程预报的局限性,提出基于卫星钟差一次差值的小波神经网络预报模型。对在轨卫星钟差求取一次差值的基础上,运用小波神经网络模型预报GPS卫星钟差,同时与GM(1,1)模型预报的结果进行比较。得出BlockΠA Cs短期预报的精度能达到0.690ns,14d预报的精度最差时依然优于1ns;其余稳定性良好的卫星钟,一天预报的结果均要优于0.207ns,预报14d卫星钟差的平均精度优于0.183ns,部分卫星钟差预报精度可以达到0.050ns,预报得到的结果可以达到GPS对实时精密单点定位的要求。
|
| 关 键 词： | 单差 卫星钟差 预报 小波神经网络 GM(1 1)模型 |
Wavelet neural network prediction for satellite in single difference |
| |
| XIONG Hongwei,CHENG Xinwen,ZHANG Haitao,MA Hairong.Wavelet neural network prediction for satellite in single difference[J].Science of Surveying and Mapping,2017,42(9). |
| |
| Authors: | XIONG Hongwei CHENG Xinwen ZHANG Haitao MA Hairong |
| |
| Abstract: | In order to deal with the limitation of current prediction model to forecast nonstationary clock offset sequence,this paper propose a model based on the first difference to predict it.On the basis of the first difference of the clock error of orbit satellite,adopt the wavelet neural network model to predict the GPS satellite clock offset,and then compared the results with the GM(1,1)model.It reveals that the prediction accuracy of Block ⅡA Cs can reach 0.690 ns in short time,the accuracy of 14 days is still better than 1 ns;The precision of the great stability satellite clock offset is better than 0.207 ns in one day,and the average accuracy of 14 days better than 0.183 ns.Part of the prediction accuracy can achieve 0.050 ns which can meet the requirements of GPS real-time precise point positioning. |
| |
| Keywords: | single difference satellite clock bias prediction wavelet neural network GM(1 1)model |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
