About One Approach to Fault-Tolerant System Modeling

Abstract

The article deals with complex control systems which are operated under severe environment conditions. One of the main tasks of development of such systems is to ensure their uninterrupted operation during a specified time interval. This task can be solved only if there is an adequate model of reliability parameters evaluation. It is suggested to simulate the reliability of complex systems using parallel semi-Markov process. An ordinary process simulates the life cycle of individual equipment unit. To simulate faults-recoveries of the system as a whole the concept of complex semi-Markov process consisting of several ordinary processes is introduced. The complex semi-Markov process is characterized by complex semi-Markov matrix, which is formed as a Cartesian product of semi-Markov matrices of ordinary processes. The semi-Markov matrices Cartesian multiplication algorithm is developed. Dependences for calculation of time intervals and probabilities of wandering through ordinary semi-Markov processes are obtained. A fast recursive method of estimation of reliability parameters is suggested. The theoretical results are confirmed by the practical calculation of the reliability indicators of the set of equipment, each of which can be in working or repairing state, as well as the reliability indicators of the system with passive redundancy.