Fractal principles of multidimensional data structurization for real-time pulse system dynamics forecasting and identification

Pulse energy conversion systems are characterized by the potential possibility of operational motion stability loss during fractions of seconds because of control high frequency effect and dynamics nonlinearity of these systems. In the case of a priori information absence, identification and forecasting in real-time mode become practically impossible. On the other hand, a priori information, mapped in the phase and parametric spaces, does not allow one-to-one decision making about the present system state even in the case of small (2-D) dimensionality of these spaces, because of impossibility of their one-to-one treatment. The principles of dynamics data structurization, based on the use of fractal regularities, are offered in the paper. These principles are general for systems with different dimensionalities, moreover structurization is realized independently for every motion type considered. As a result, 2-D projections of the special spaces are formed in which one-to-one treatment of the acting system state becomes possible in the real-time mode. As an example 4-D mathematical model of pulse system and random variation of 2-D parametric vector are considered.