20Jan 2019

VERSATILITY OF TIME-VARYING AMPLITUDE METHOD IN HARMONIC ANALYSIS OF DISCRETE DATA.

  • Faculty of Science,Nigerian Defence Academy, Kaduna.
Crossref Cited-by Linking logo

  • Abstract
  • Keywords
  • References
  • Cite This Article as
  • Corresponding Author

This paper proposes a method that overcomes the problem of the amplitude not being zero at the starting time .when the Cosine waveform is used in Harmonic analysis. The paper employs the method of time-varying amplitude method to get a satisfactory answer to this nagging issue of the amplitude having a value different than zero when the motion has not started for discrete data. Models obtained using the time-varying amplitude method have shown to produce better results than those from the constant amplitude model when the traditional time-invariant Fourier method is employed in Fourier analysis. The method is also flexible in the sense that it can be used to convert a decaying amplitude to a growing one and vice versa and at the same time produces amplitudes that will match oscillations whose amplitudes change with time. Using the time-varying amplitude method reduced the Sum of squares of errors of the constant amplitude model by about 56% and increased the amplitude by over 91 %, and more weights are giving to the recent data like in the smoothing methods .Besides, the method has reduced the effect of spurious correlation in time series.

  1. Baghazou,Y, Y, R.E Morrison, P.F. Ribeiro and C.A. Duque (2001); Time-varying Waveforms/Distortions Power Systems. John Wiley and Sons
  2. Bird John (2000) Higher Engineering Mathematics.Newness,An Imprint of Elsevier
  3. Bloomfield, P.(1998);Fourier Analysis of Time Series .Second Edition. I Wiley and Sons Inc.
  4. Bowerman,B.L,Koehler A.B and Pack D.J(1990) Forecasting Time Series with increasing Seasonal Variations, Journal of Forecasting,No.5,Vol9,419-436,1990
  5. Box,G.E.P., and G.M, Jenkins(1976) ; Time Series Forecasting and Control.EnglewoodCliffs,NJ:Printice Hall
  6. Chris Hintz (2017); Number Cruncher Statistical software(NCSS12)
  7. Gardner, A. William(1998); Statistical Spectral Analysis. Printice Hall
  8. Okafor, U.L and Oladejo,M.O (2013);?Another Look at Fourier Coefficients, Forecasting Nonstationary Time Series? Using Time-Dependent Fourier Coefficients.? Journal of Computing,Communication and Instrumentation Engineering (IJCCIE),Vol.1,Issue1 (2014), pp186-92
  9. L,Okafor (2015);Discrete Approximation of Nonstationary Sequences by Time-Weighted Fourier Amplitude. PhD Thesis, Nigerian DefenceAcademy,Kaduna,Nigeria
  10. V,Ossiotzko (2005); Fourier Analysis Module. Introductory Laboratory Course Physcis.University of Oldenber
  11. Yang Cheng(1997);Bayesian Time Series: Financial Models and Spectral Analysis,PhDDissertation.Duke University.
  12. U. Yule (1926);Why Do We Sometimes Get Nonse Correlation Between Time Series?. A Study in Sampling and the Nature of Time Series.Journal of Royal Statistical Society.Vol 89,1926,pp1-44.

[Uchenwa linusokafor, O.Oladejo, D.T.Chinyo and C.O.Uwa. (2019); VERSATILITY OF TIME-VARYING AMPLITUDE METHOD IN HARMONIC ANALYSIS OF DISCRETE DATA. Int. J. of Adv. Res. 7 (Jan). 1037-1040] (ISSN 2320-5407). www.journalijar.com

UCHENWA LINUS OKAFOR
NIGERIAN DEFENCE ACADEMY,KADUNA

DOI:

Article DOI: 10.21474/IJAR01/8419      
DOI URL: https://dx.doi.org/10.21474/IJAR01/8419

Download Full Paper

Download PDF No. of Downloads: 6 | No. of Views: 48