A Number of Issues about Converting GPS Height by BP Neural Network

Author

Lu Tieding ; Zhou Shijian ; Chen Xijiang

Author_Institution

East China Inst. of Technol., Fuzhou, China

fYear

2010

fDate

23-25 April 2010

Firstpage

1

Lastpage

4

Abstract

To convert GPS height by Artificial Neural Networks, there are still many problems which need further research. In this paper we conduct more in-depth analysis and study more about the specific problems such as the determination of hidden layer, the determination of hidden layer nodes, network training times, selections of learning rate and initial weight values and so on. After the improved setting of the parameters , we can make the improved BP Neural Network model for the conversion of GPS height. Then some useful conclusions are obtained.

Keywords

Global Positioning System; backpropagation; neural nets; telecommunication computing; BP neural network; GPS height conversion; Global Positioning System; hidden layer nodes; initial weight values; learning rate selection; network training times; Algorithm design and analysis; Artificial neural networks; Geoscience; Global Positioning System; Measurement standards; Measurement techniques; Neural networks; Testing; Transfer functions; Uncertainty;

fLanguage

English

Publisher

ieee

Conference_Titel

Biomedical Engineering and Computer Science (ICBECS), 2010 International Conference on

Conference_Location

Wuhan

Print_ISBN

978-1-4244-5315-3

Type

conf

DOI

10.1109/ICBECS.2010.5462426

Filename

5462426