Research On Reliability Measurement Of Industrial Control System With Markov Reward Model

As the core of national critical infrastructure,industrial control systems are widely used in transportation,electric power,petroleum and other industries.In recent years,with the introduction of "Made in China 2025",the reliability of industrial control systems has also attracted great attention from academia and enterprises.In this thesis,we use the multi-state theory to construct the reliability measurement model of industrial control network system.The basic assumption of traditional industrial control system reliability assessment research is the two-state assumption,i.e.,there are only two states of normal operation and complete failure,and the two-state assumption does not reflect the real essential characteristics of industrial network control system when portraying it.This thesis defines the reliability of an industrial control system as the probability of its ability to complete tasks.Through the top-down analysis of the industrial control system's architecture,the industrial control system's architecture model is obtained.Component-level,subsystem-level and system-level reliability are calculated from bottom to top.By introducing a Universal Generating Function,the reliability of the subsystem is calculated from the reliability of the component level,and then the reliability of the entire industrial control system is calculated from the reliability of the subsystem.It solves the problem of level division of reliability calculation of complex industrial control system,and decomposes the complex calculation problem of industrial control system reliability measurement into several simple calculation sub-problems to solve the problem of "state explosion" of industrial control system.In the reliability calculations at different levels,the thesis verifies the proposed related calculation methods through actual cases,and proves the accuracy and feasibility of the proposed methods.The method proposed in the paper provides a solid theoretical foundation for the development of reliability measurement software for modern complex industrial control systems.