Fractal analysis of distillation unit time series in prediction and control problems

Abstract

The behavior of distillation unit as a control object is characterized by stochasticand chaotic manifestations. This requires their identification by the nonlineardynamics methods for organizing the specific respective control strategies. Weused the synergetic methods and the theory of deterministic chaos for solving problems of distillation unit prediction and control. Time series analysis wasconducted using the Rescaled Range algorithm of the analysis of Hurst, the coatingmethod and the correlation methods. Time series are processed using the software package FRACTAN.