基于灰色BP网络的卫星钟差预报  

Satellites clock bias forecast based on gray system and BP neural network

作　　者：郭振华[1] 韩保民[1] 赵金生[2] 

机构地区：[1]山东理工大学建筑工程学院,山东淄博255091 [2]淄博职业学院建筑工程学院,山东淄博255314

出　　处：《山东理工大学学报（自然科学版）》2012年第4期39-42,共4页Journal of Shandong University of Technology:Natural Science Edition

基　　金：国家自然科学基金资助项目(41074001);山东省高等学校科技计划项目(J09LE09)

摘　　要：精密卫星钟差是精密定轨的重要因素之一.将BP神经网络模型引入卫星钟差预报,分别对少量及大量的卫星钟差数据进行预报,将其预报精度和灰色模型GM(1,1)预报结果进行分析比较,证明BP神经网络模型适于卫星钟差预报.通过对IGU超快速星历中的卫星钟差进行预报,发现前期预报精度很高,验证了其实际有效性.The precise satellite clock bias is one of the most important fneural netwoactors for satellite orbit determination. In this paper, the BP rk model was used for satellite clock bias forecast, using a vast amount of data and a few data respectively. Then the forecast accuracy of BP model was compared with that of gray model GM （1, 1）. The results suggest that the BP neural network model is suitable for the satellite clock bias forecast. Further more, the BP neural network model was used for forecast based on ultra-rapid clock bias data, and the high precision in prophase indicated it is effective.

关 键 词：BP神经网络 灰色模型GM(1  1) 卫星钟差 预报 

分 类 号：P228[天文地球—大地测量学与测量工程]