Polonsky O. Synthesis of classifiers of control objects' states under conditions of indefiniteness

The thesis is devoted to the development of methods of synthesis of new adaptive observers - invariable classifiers of control objects' states under conditions of indefiniteness. In the central part of the work, the significant attention is paid to the optimization of classifiers' structures in the distribution area of observations ranks. On the basis of numerical examples it has been proved the efficiency of methods and recommendations revealed in this thesis. Results of the work undertaken can be applied as the basis to build the control systems with the usage of the complicated technical objects under conditions of structural and parametric indefiniteness. The accuracy of results is proved by applications.