STOCHASTIC MECHANISMS OF THE ACTIVE SYSTEMS FUNCTIONING

Abstract

The stochastic control of large-scale hierarchical systems implies the consideration of a special type of human factor-elements activity connected with the availability of the elements own goals. The author consider the situation in which a prescient active element may predict the centre controlling actions and choose its states in such a way that its effect on the results of state estimation and stochastic control will maximize its own goal function. They analyse the problems of the synthesis of stochastic functioning mechanisms (including estimation, planning, control and stimulation procedures) in the hierarchical, active systems, with the information, obtained from the active elements in the course of their functioning, being used for estimating their states and changing the parameters of the mechanism proper and reaching the aim of the centre. The author present the statements of the optimal synthesis problems. The constructive sufficient conditions for the existence of stochastic mechanisms were found with the typified procedures of state estimation in the process of the active systems control. The problem of developing and implementing the stochastic mechanisms in the control practice of large-scale social and economical systems are analysed.