Predicting the Polish timescale UTC(PL) based on the corrections designated by the UTC and UTCr scale

Author

Sobolewski, Lukasz

Author_Institution

Inst. of Electr. Metrol., Univ. of Zielona Gora, Zielona Góra, Poland

fYear

2013

fDate

21-25 July 2013

Firstpage

658

Lastpage

661

Abstract

The article presents the results of research on the prediction of the Polish timescale UTC(PL) based on the corrections designated according to UTC and UTCr scale. In both cases for prediction the GMDH neural networks and time series analysis method were used. For predicting the UTC(PL) the influence of the input data preparation method for the GMDH neural network was examined. The obtained results were compared with the predictions designated using linear regression method. Better results of predicting the UTC(PL) were obtained using GMDH neural network. The most preferred results have been obtained for the input data containing only the corrections designated by the UTCr scale.

Keywords

atomic clocks; neural nets; regression analysis; time series; GMDH neural networks; Polish timescale prediction; UTC scale; UTC(PL) prediction; UTCr scale; linear regression method; time series analysis method; Biological neural networks; Linear regression; Market research; Time series analysis; Training; Training data; GMDH neural networks; Polish timescale UTC(PL); Rapid UTC; atomic clock;

fLanguage

English

Publisher

ieee

Conference_Titel

European Frequency and Time Forum & International Frequency Control Symposium (EFTF/IFC), 2013 Joint

Conference_Location

Prague

Type

conf

DOI

10.1109/EFTF-IFC.2013.6702099

Filename

6702099