DEFINING THE SAFETY INDICATORS OF A UNRECOVERY FAULT-TOLERANT SAFETY CRITICAL SYSTEM BY THE STATE SPACE METHOD

Safety Critical Systems (SCS) play a key role in critical areas of activity where high safety, reliability and continuity of operations are required. Such systems include military, space, energy, aviation, medical and transportation complexes that operate in extreme conditions and must perform their functions regardless of external and internal influences. The main characteristics of such systems are the ability to perform their functions in case of loss of operability of their subsystems or modules, the ability to adapt to changing operating conditions and a high level of protection against external threats. The object of study in this paper is a unrecovery fault-tolerant safety critical system with a majority structure. Such a system consists of an odd number of modules of the same type and a majority element (vote system). The subject of the study is the safety indicators of a fault-tolerant system of responsible assignment. The article presents a methodology for determining the safety indicators of a safety critical system using the state space method. This method, in contrast to the methods of fault trees, dynamic fault trees, event trees, and FMEA/FMECA analysis, allows taking into account the influence of the fault-tolerant system behavior algorithm on the occurrence of emergencies and, as a result, obtaining reliable quantitative values of safety indicators. The peculiarity of the state space method is that safety and reliability indicators are determined from a single model. This makes it possible to determine the impact of the configuration of the structure of a fault-tolerant system on its safety indicators, which is impossible when using other methods of safety analysis. The methodology is illustrated on a specific example of a unrecovery fault-tolerant safety critical system. The method of fault tree analysis is used to validate the results of the study. The developed methodology for generating safety indicators from the subspace of inoperable states makes it possible to obtain accident functions of vibration-resistant systems, which are time dependencies of the probability of occurrence of minimum severity. In contrast to the minimum severity functions, the accident functions allow taking into account the impact on safety indicators of the peculiarities of the fault-tolerant system behavior algorithm in case of performance disruptions.