Sukhomlyn R. Mathematical and information support for the process of determining disorders in technical systems

The thesis is devoted to solving actual scientific tasks improving of methods and developing models and computer programs for early diagnosis of technical systems nonstationary processes disorders. As the technical objects we considered electric motor, reduction gear of turbine engine and gas-jet mill. ARIMA and FARIMA models were built for viewed technical systems, on their basis we developed the algorithm of disorders detection. We developed software for detection process disorders in technical systems (electric motor, aviation reduction gear, gas-jet mill and for expansion and universalization of the use of developed models and methods - electrophysiological system of heart) based on the methods of wavelet and Hilbert-Huang transform. We proved the adequacy of these methods and models in the analysis of the problem of nonstationary and noisy signals, making it possible to detect disorders at an early stage of their development.