中文核心期刊
CNKI期刊全文数据库
中国科学引文数据库(CSCD)源期刊
中国科技论文统计源期刊
万方数据知识服务平台
英国《科学文摘》(SA)
美国化学文摘(CA)
俄罗斯《文摘杂志》(AJ)
德国《天文学与天体物理学文摘》(AAA)
英国《中国天文学和天体物理学》(SCI收录)全文摘译期刊之一
《中国学术期刊文摘》
《中国物理文摘》
《中国天文学文摘》

Chinese Journal of Space Science ›› 2016, Vol. 36 ›› Issue (1): 83-91.doi: 10.11728/cjss2016.01.083

Previous Articles     Next Articles

High-precision orbit prediction for high-altitude orbit satellite based on ANN model

HUANG Jin1, ZHANG Zhengqiang1, ZHANG Yuzhe1, YANG Ge1, LI Xiaojie2   

  1. 1. Beijing Remote Sensing Institute, Beijing 100192;
    2. Beijing Satellite Navigation Center, Beijing 100094
  • Received:2014-03-10 Revised:2014-09-03 Online:2016-01-15 Published:2016-01-06

Abstract:

As to the problem of rapid attenuation of predictive orbit precision obtained by dynamics model, a new method of satellite orbit prediction based on Artificial Neural Network (ANN) model was proposed. By using ANN as the tool of constituting orbit prediction model, a training sample set was built based on the orbital dynamics characters, the best training sample was searched based on the character of current orbital error, and ANN model was got through training to compensate current orbit. The experimental results showed that, the effectiveness of the improvements varies with different satellites and initial epochs. The error of orbit predictions for the 4-day prediction was reduced from 43m to 15m; and for the 8-day prediction, it was reduced from 183m to 80m. The improvement rates of predicting 4-day and 8-day were 78.33%, 88.33% respectively.

Key words: Geostationary Orbit satellites (GEO), Inclined Geostationary Orbit Satellites (IGSO), Artificial neural network, Orbit prediction, Digital filtering, Smoothing

CLC Number: