THE APPLICATION OF DIFFUSIONAL TECHNIQUES IN TIME-SERIES ANALYSIS TO IDENTIFY COMPLEX FLUID DYNAMIC REGIMES
Massimiliano Giona, Alessandro Paglianti, Alfredo Soldati · Fractals · 1994
The diffusional analysis of time series is introduced and compared with corresponding results derived from rescaled-range (R/S) techniques. It is shown by means of simple examples that diffusional analysis is able to capture some details of the statistics of the time series which can become lost when other techniques (e.g. R/S) are applied. Examples of the application of this method to chaotic time series are briefly discussed. The problem of the identification of fluid dynamic regimes is investigated in detail by considering experimental time series derived from two-phase flow measurement in tubes. The results obtained confirm the validity of this method in dealing with time series derived from periodic or quasiperiodic phenomena with superimposed stochastic perturbations.