Tehničko veleučilište u Zagrebu · Zagreb

Algebraic Decision Diagrams in Multi-State Reliability Analysis of Nuclear Power Plant Systems

izvorni znanstveni rad

izvorni znanstveni rad

Algebraic Decision Diagrams in Multi-State Reliability Analysis of Nuclear Power Plant Systems

Vrsta prilog sa skupa (u zborniku)
Tip izvorni znanstveni rad
Godina 2024
Nadređena publikacija Proceedings of the 14th International Conference of the Croatian Nuclear Society
Stranice 122, 10
Status objavljeno

Sažetak

The reliability analysis which takes in account multiple possible states of components is considered as a multi-state reliability analysis. Although multi-state reliability models provide more realistic and more precise representations of Nuclear Power Plant systems, their analysis is considered far more complex and demanding. A popular way to approach such an analysis consists in the application of stochastic processes, especially Markov chains. Unfortunately, the number of transitional state spaces in Markov chains grows exponentially with the increase in the number of component states. Algebraic Decision Diagrams prove to be an efficient data structure for storing Markov chains and can be used to perform such an analysis. Preliminary results on smaller models show the applicability of algebraic decision diagrams to represent Markov chains describing hundreds or even thousands of states from Nuclear Power Plant systems with multi-state components. These results justify the implementation of algebraic decision diagrams techniques to automate Markov chain analysis of systems with extremely large state space models. Experimental results from a prototype of an algebraic decision diagrams-based analysis of a simplified model of the Nuclear Component Cooling Water system are provided to substantiate conclusions.

Ključne riječi

Probabilistic Safety Assessment (PSA), Dynamic Fault Tree (DFT), Stochastic Process (SP), Algebraic Decision Diagrams (ADD), Multi-state Systems (MSS)