Fomin O. Information technology of secondary spectral and statistical identification of nonlinear inertial objects in technical diagnostic systems

The dissertation is devoted to solving the problem of increasing the reliability and efficiency of diagnosing based on information technology of secondary spectral and statistical identification of continuous nonlinear inertial objects of various nature in systems of technical diagnostics. Based on the proposed approach, methods of nonparametric identification of nonlinear inertial objects of the "black box" type in the form of multidimensional transitive functions are developed and investigated. The family of diagnostic models of nonlinear inertial objects is proposed on the basis of spectral transformations of cross sections of multidimensional weight functions for compression of the space of features and statistical methods machine learning models in received diagnostic spaces. Based on these models and methods an information technology of the secondary spectral-statistical identification of nonlinear inertial objects is developed on the basis of integral nonparametric dynamic models in the conditions of incomplete a priori information. The obtained results are brought to practical realization in the form of instrumental means of the automated system of technical diagnostics of nonlinear inertial objects of various nature.