Reliability Analysis of Cyber-Physical Systems

Abstract

A formalized representation of cybernetic and physical components for cyber-physical systems (CPS) is proposed. It matches numerical procedures' design for operational reliability analysis of multi-component industrial systems. The reliability indices are supposed to be estimated for the systems subject to intensive industrial exploitation. The viability for CPS is executed through the artificial neural networks (ANN s) mathematical model application. For this purpose the next outputs of the integrated model CPS+ ANN are calculated: the equipment conflict rate factor, the information completeness factor, the reliability index and the operator activity coefficient which is an important indicator for the system's ability to function in a fully automatic mode. The ANNs tuning utilizes the information on the states of physical components and cybernetic components via accessible historical data and experts' knowledge treatment. A digital electrical substation (DSS) is discussed as a demonstration example of the CPS. The reviewed example allows one to clearly present the proposed approach and algorithm.