Industrial Automatic Control Systems and Controllers Annotation << Back Algorithms for Estimating and Optimizing the Parameters of Information Subsystems of Guidance Complexes based on their Division Into Conditionally Dependent Classes Gromov Yu.Yu., Bunin A.V., Potapov A.N., Nachalov A.L., Kyznechsov V.A. The issues of estimating the parameters of information subsystems with abrupt changes in structure with unknown distribution parameters are considered. It is shown that “expanding” the state vector to clarify the a priori description of the process leads to an increase in computational costs, and the resulting estimation accuracy decreases due to the a priori uncertainty of transition intensities. The quality of identifi cation is higher, the greater the differences in the intensities of state changes for the identifi ed structures, the greater the difference in the coordinates of the identifi ed sources of information; less noise level of the meter. The developed algorithms for estimating and optimizing the parameters of information subsystems of active homing complexes based on their division into conditionally dependent classes made it possible to prove that increasing uncertainty about the parameters of abruptly changing processes can be achieved using complex control laws. but the “increase” of the a priori complexity of the mixed process is limited by the level of observation noise. There are restrictions on the degree of complexity of the process, and at a certain value, increasing complexity does not lead to an increase in resource. Increasing the level of uncertainty is achieved by randomizing the structure of the system. The use of uncertainty about the number of states leads to a decrease in the system’s ability to estimate the parameters of the opposing system, despite the optimality of the decision rules. The high accuracy of the a priori forecast makes it possible to use the developed procedure for estimating the parameters of information subsystems of guidance complexes based on their division into conditionally dependent classes to construct algorithms for adaptive control of dynamic information systems. Keywords: structure, stability, assessment, parameter, information subsystem, vector, information, algorithm, filtering. Pp. 37-47.