| Abstrakt: |
he reliability and serviceability of metal structures depends on the quality of control over the technical condition and mechanical properties of materials in accordance with international standards. However, in the course of operation, deviations from the normative values of material properties occur due to the uncertain nature and magnitude of the loads, which necessitates periodic shutdowns of the equipment for the purpose of diagnostics. Uninterrupted operation of elements of power equipment of marine vessels depends on the quality of control of their technical condition by physical diagnostic methods. The most effective method of reducing operating costs and increasing the reliability of equipment is maintenance based on interactive monitoring of its condition, detection of malfunctions and forecasting of energy equipment parameters. This makes the tasks of control, diagnosis and forecasting of energy equipment parameters particularly relevant. At the same time, the use of various diagnostic methods does not allow taking into account all the features of real operating conditions. This problem is especially acute in conditions of various extreme situations and peak loads, which cannot be taken into account. Probabilistic methods, in particular Markov chains, are a promising area of research. The system of intelligent diagnostics and monitoring of ship power plant turbochargers using Markov chains is presented. The novelty of the developed methodology lies in replacing discrete time intervals of the diagnostic process with a sequence of states of technical objects. In this formulation, Markov chains represent a synthetic property that accumulates diverse factors. Randomization of stochastic diagnostic and monitoring processes of ship power plant components enables an increase in reliability of equipment under severe operating conditions. The results of calculations of digitalization of experimental data, calculations of transition matrices and construction of an orgraph allowing to study the kinetics of damage accumulation in real time are presented. [ABSTRACT FROM AUTHOR] |
