prediction method of low earth orbit based on bp neuron network
- Paper number
IAC-08.C1.3.8
- Author
Dr. Xiaoning Du, Xi'an Jiaotong University, China
- Coauthor
Mr. Zhibing Sun, Xi'an Satellite Control Center, China
- Coauthor
Prof. Yongxuan Huang, Xi'an Jiaotong University, China
- Coauthor
Prof. Jisheng Li, Xi'an Jiaotong University, China
- Coauthor
Dr. Xiangyang Yu, Xi'an Jiaotong University, China
- Coauthor
Ms. Jingjiang He, Xi'an Jiaotong University, China
- Year
2008
- Abstract
The change of atmospheric density which is caused by the solar activity and geomagnetic activity is an important factor to influence the change of low earth orbit. How to make exactly use of the change of solar activity and geomagnetic activity for improving the satellite orbit calculation, orbit prediction and orbit control, it will be an important subject and technological challenge, because it is a very difficult work for the accurate prediction of the solar activity and geomagnetic activity, moreover, there is not any accurate model to describe how the atmospheric density varies with the solar activity and geomagnetic activity. The existing spacecraft orbit prediction models all were carried on based on the parameters of atmospheric density of the classical model. Since the parameters of atmospheric density cannot be chosen in real-time, the exact prediction and the exact control for low earth orbit are very difficultly carried on in real orbit calculation. We present in this paper a prediction method of low earth orbit which is based on BP neuron network. Regard the parameters of solar activity and geomagnetic activity as the input parameters of BP neuron network, and the parameters of satellite orbit as the output parameters of network to train the network, thus, a BP neuron network of low earth orbit is obtained, at the same time, the orbit prediction can be performed. Finally, an example is provided to demonstrate the exactness of this method in the paper.
- Abstract document
- Manuscript document
(absent)