V. V. Dubarenko, V. G. Kurbanov

This article observes method of solving the problems of control upon dynamic objects. Originality is in the fact that optimization algorithms are presented in the form of ultimate automata and then taken to the form of linear sequential machines which allow the numerical techniques to define measures of computing and combinatorial complexity of control algorithms by means of [0,1] matrices analysis. Search algorithms of optimization can be presented in the form of two parts: arithmetic part where the vector of dynamic object condition is calculated, and logical part where the vector of logic variables necessary for decision-making on control impact valuation is specified. Valuations of logic variables attributes as feedback are brought to the arithmetic part for making the next iteration. The process continues till the valuations of logic variables attributes satisfy given conditions. While building the knowledge data bases of logic part linguistic link of conditionals is used, e.g. 'if..., than...' relevant to the logic operation - implication.