Enhancing of Technical Systems Reliability by Implementing of Risk-Oriented Diagnostics

This paper discusses issues dedicated to increasing of complicated technical systems' reliability in terms of survivability assurance. The classification of diagnostic varieties is proposed, with the allocation of a functional diagnostics branch directed at system risks identification. It is shown that in order to ensure the survivability, it is necessary to calculate the probability of risks occurrence, which could be caused by factors of different nature, and to determine critical risks that can lead to the survivability violation. An example of the risk probability calculating for a system with complicated configuration, in which interactions between the system elements occur in pairs, crosswise, and recurrently, using the theory of the discrete Markov chains, is provided. Method for calculating of the survivability function is proposed, and a probability model of a risk event occurrence is compiled, which can be used to manage risks in technical systems. An algorithm for the regression equations constructing for probabilities of various risk events types is shown, that allows identifying the most significant risks' components, aimed on further minimization of risks by using a decision support system.