Comparison of linear regression method and GMDH neural network in predicting the UTC(PL) national timescale

• Autorzy: Sobolewski Ł.

Identyfikatory

DOI

10.15199/48.2018.11.01

Warianty tytułu

PL

Porównanie metody regresji liniowej i sieci neuronowych GMDH w prognozowaniu krajowej skali Czasu UTC(PL)

• Języki publikacji: EN

Abstrakty

EN

The paper presents research results on predicting the Polish Timescale UTC(PL) by the means of GMDH-type neural network and linear regression method for data prepared in the form of time series built on the basis of [UTC - UTC(PL)] and [UTCr - UTC(PL)] deviations and values of a phase time from UTC(PL). The obtained results show comparable prediction quality by means of GMDH-type neural network with prepared procedure of predicting and linear regression method modified by the author for timescale characterized with high stability.

PL

W pracy przedstawiono wyniki badań nad prognozowaniem Polskiej Skali Czasu UTC(PL) przy zastosowaniu sieci neuronowej typu GMDH oraz metody regresji liniowej dla danych przygotowanych w formie szeregu czasowego, zbudowanego z wartości odchyleń [UTC - UTC(PL)] oraz [UTCr - UTC(PL)] oraz wartości czasu fazowego z UTC(PL). Wyniki badań pokazały porównywalną jakość prognozowania z zastosowaniem sieci neuronowej typu GMDH i opracowanej procedury prognozowania oraz zmodyfikowanej przez autora metody regresji liniowej w przypadku skali czasu charakteryzującej się dużą stabilnością.

Słowa kluczowe

EN

UTC(k) timescale; atomic clock; predicting [UTC – UTC(k)]; GMDH neural network; linear regression method

PL

skala czasu UTC(k); zegar atomowy; prognozowanie [UTC-UTC(k)]; sieci neuronowe GMDH; metoda regresji liniowej

• Wydawca: Wydawnictwo SIGMA-NOT
• Czasopismo: Przegląd Elektrotechniczny
• Rocznik: 2018
• Tom: R. 94, nr 11
• Strony: 1--5
• Opis fizyczny: Bibliogr. 26 poz., rys., tab.

Twórcy

autor

Sobolewski Ł.

• Uniwersytet Zielonogórski, Instytut Metrologii, Elektroniki i Informatyki, ul. prof. Z. Szafrana 2 65-516 Zielona Góra

Bibliografia

• [1] BIPM Annual Report on Time Activities, vol. 11, Sevres BIPM, 2016
• [2] Petit G., Arias F., Harmegnies A., Panfilo G., Tisserand L., UTCr: a rapid realization of UTC, Metrologia, 51 (2014), 33-39
• [3] Bernier L. G., Use of the Allan Deviation and Linear Prediction for the Determination of the Uncertainty on Time Calibrations Against Predicted Timescales, IEEE Transactions on Instrumentation and Measurement, 52 (2003), 483-486
• [4] Czubla A., Konopka J., Nawrocki J., 2006 Realization of atomic SI second definition in context UTC(PL) and TA(PL), Metrology and Measurement Systems, 2 (2006), 149-159
• [5] Davis J. A., Shemar S. L., Whibberley P. B., A Kalman filter UTC(k) prediction and steering algorithm, Proc. Joint IEEE FCS EFTF (San Francisco, USA, 2-5 May 2011), 779-784
• [6] Panfilo G., Tavella P., Atomic clock prediction based on stochastic differential equations, Metrologia, 45 (2008), 108-116
• [7] Kaczmarek J., Miczulsk i W. , Kozioł M., Czubla A., Integrated system for monitoring and control of the national time and frequency standard, IEEE Transactions on Instrumentation and Measurement, 62 (2013), 2828-2838
• [8] Liao C. S., Chu F. D., Lin H. T., Tu K. Y., Chung Y. W., Hsu W. C., Formation of a paper neural-fuzzy time scale in the Eastern Asia, Joint IEEE FCS EFTF (San Francisco, USA, 2-5 May 2011), 292-295
• [9] Rovena G. D., Abgrall M., Chupin B., Guena J., et al., The New UTC(OP) Based on LNE-SYRTE Atomic Fountains, Proc. Joint UFFC, EFTF and PFM (Prague, Czech Republic, 21-25 July 2013), 649-651
• [10] Abgrall M., Bize S., Chupin B., et al., Performances of UTC(OP) based on LNE-SYRTE atomic fountains, Proc. Eur. Frequency Time Forum (Neuchatel, Switzerland, 23-26 June 2014), 564-567
• [11] Norgard M., Ravn O., Poulsen N., Hansen L., Networks for Modelling and Control of Dynamic Systems, Springer Verlag, 2000
• [12] Tadeusiewicz R., About usefulness of neural networks in electrical engineering problems, Electrical Review, 2 (2009), 200-211
• [13] Korbicz J., Artificial neural networks and their application in electrical and power engineering, Electrical Review, 9 (2009), 194-200
• [14] Nelles O., Nonlinear System Identification. From Classical Approaches to Neural Networks and Fuzzy Models, Springer Verlag, 2001
• [15] Luzar M., Sobolewski Ł., Miczulsk i W., Korbicz J., Prediction of corrections for the Polish time scale UTC(PL) using artificial neural networks, Bulletin of the Polish Academy of Sciences: Technical Sciences, 61 (2013), 589 -594
• [16] Miczulski W., Sobolewski Ł., Influence of the GMDH neural network data preparation method on UTC(PL) correction prediction results, Metrology and Measurement Systems, 19 (2012), 123-132
• [17] Sobolewski Ł., Predicting the corrections for the Polish Timescale UTC(PL) using GMDH and GRNN neural networks, Proc. Eur. Frequency Time Forum (Neuchatel, Switzerland, 23-26 June 2014), 475-478
• [18] Onwubolu G., GMDH - Methodology and Implementation in C, Imperial College Press, 2015
• [19] Farlow S. J., Self-organizing Methods in Modelling: GMDHtype Algorithms, Marcel Dekker, 1984
• [20] Sobolewski Ł., Predicting the Polish timescale UTC(PL) based on the corrections designated by the UTC and UTCr scale, Proc. Joint UFFC EFTF and PFM (Prague, Czech Republic, 21-25 July 2013), 658-661
• [21] Sobolewski Ł., Miczulski W., Application of neural networks for predicting selected time scales on the basis of UTC and UTCr scales, Electrical Review, 10 (2016), 258-261
• [22] Sobolewski Ł., Application of the neural networks for predicting the corrections for the national timescale UTC(PL), University of Zielona Gora Press, 2016
• [23] Miczulski W., Sobolewski Ł., Algorithm for predicting [UTC - UTC(k)] by means of neural networks, IEEE Transactions on Instrumentation and Measurement, 66 (2017), 2136-2142
• [24] Sobolewski Ł., Application of GMDH type neural network for predicting UTC(k) timescales realized on the basis of hydrogen masers, Proc. Joint Conference of the European Frequency and Time Forum and IEEE International Frequency Control Symposium (Besançon, France, 10-13 July 2017), 42-46
• [25] Sobolewski Ł., Predicting the Lithuanian Timescale UTC(LT) by means of GMDH neural network, Bulletin of the Military University of Technology, 66 (2017), 31-41
• [26] Caldwell R. B., Performance metrics for neural networkbased trading system development, NeuroVest Journal, 3 (1995), 22-26

Uwagi

Opracowanie rekordu w ramach umowy 509/P-DUN/2018 ze środków MNiSW przeznaczonych na działalność upowszechniającą naukę (2018).