par Cornélissen, Germaine;De Prins, Jean
Référence Journal of Interdisciplinary Cycle Research, 6, 1, page (89-94)
Publication Publié, 1975
Article révisé par les pairs
Résumé : It has been shown that the chosen numerical integration method corresponds to a realistic view of data processing. Sampling conditions and frequency range have been carefully examined to define the Fourier coefficients. The estimation of the errors has pointed at the ability of the spectral analysis method. The definition of the first order autocovariance coefficient has led to the knowledge of the time dependence of both amplitude and frequency.