Ponedilok V. Methods and tools to construction and realization integro-degree models of processes restoration input signals nonlinear dynamic objects

In the dissertation work the scientific and technical problem of use and development methods of mathematical and computer modelling of nonlinear dynamic objects presented by integro-degree Volterra series is considered and solved. The methods for solving Volterra bilinear integral equations of the first kind are solved with a differential regularization parameter that allows the use of integro-degree Volterra series for a wide class of tasks for the recovery of input signals at the input of nonlinear dynamic objects. The methods of identification of dynamic objects using algorithms of differentiation of experimentally obtained functional dependences, which are used for construction Volterra multidimensional kernels, are implemented, which allows to improve the stability of the computational process of identification. The scientific novelty consists in creation methods mathematical and computer modelling of nonlinear dynamic objects based on integro-degree Volterra series using regularization algorithms of numerical differentiation and solving of Volterra bilinear integral equations I-th kind, which allow to improve the solving efficiency. tasks of analysis, identification, management, etc. The practical significance of the results obtained is that the developed methods and means of constructing and researching approximation integral models of dynamic objects, as well as the set of programs, expanding the possibilities of using computer tools in the tasks of mathematical and computer modelling of nonlinear dynamic objects type "input-output" and in tasks of creation of modern systems of measurement, control, control and diagnostics.