Functional Dependability of Distributed Control of Multi-Zone Objects Under Failures Conditions

Abstract

The significance of the functional dependability problem of distributed control systems is constantly increasing due to the spread of inexpensive controllers and sensors and the creation of Internet of Things (IoT) systems. In the case of distributed control of multi-zone objects with the mutual influence of zones, the problem is complicated by changes in the operating modes of local control systems (LCS) due to the failures of elements of the systems of neighbouring zones. This, in turn, leads to a change in their failure rates. In this work, the functional dependability model of the decentralized distributed control system for multi-zone objects was developed and investigated. The model is obtained in the form of an inhomogeneous Markov chain of the system states caused by failures of elements of LCS. The probabilities of transitions between states are described by the Chapman-Kolmogorov equations and additional equations that describe the relationship between the parameters of neighbouring zones and the corresponding LCS. The impact of Data Transmission System (DTS) overload, controller failures, sensor failures, and DTS failures on the probability of the zones’ state exceeding the permissible limits was studied. The functional safety assessment model of decentralized control of multi-zone objects, proposed in the previous works of the author was used for the theoretical analysis. Verification of the theoretical model is achieved through simulation, leveraging a library of modules developed by the authors on the Scilab/Xcos platform. The simulation showed that the "cascade failures" process is probable for multi-zone objects with coordinated control according to a global criterion and a strong connection between the parameters of the zones, which leads to the rapid loss of functional dependability.