SHyFTOO, an object-oriented Monte Carlo simulation library for the modeling of Stochastic Hybrid Fault Tree Automaton. (15th May 2020)

• Record Type: Journal Article
• Authors: Chiacchio, Ferdinando; Aizpurua, Jose Ignacio; Compagno, Lucio; D'Urso, Diego
• Is Part Of: Expert systems with applications. Volume 146(2020)
• Publication Date: 2020-05-15
• Subjects: Discrete event simulation -- Multi-state systems -- Repairable systems -- Dynamic reliability -- Matlab® -- Dynamic fault tree; ARCE Average Relative Cumulated Error -- ATS Adaptive Transition Systems -- BE Basic Event -- BDMP Boolean Driven Markov Process -- DES Discrete Event Simulation -- DFT Dynamic Fault Tree -- DRBD Dynamic Reliability Block Diagram -- FTA Fault Tree Analysis -- HBE Hybrid Basic Event -- OO Object-Oriented -- pdf Probability Density Function -- RAMS Reliability, Availability, Maintenance, Safety -- RBD Reliability Block Diagram -- SFT Static Fault Tree -- SHA Stochastic Hybrid Automaton -- SHyFTA Stochastic Hybrid Fault Tree Automaton -- SPDEs Stochastic Partial Differential Equations -- SPN Stochastic Petri Nets; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2019.113139
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: State of the art of main academic tools for the dependability assessment. Object oriented software architecture for the development of a dependability tool. Matlab® toolbox library to model and simulate Stochastic Hybrid Fault Tree Automatons. Case study characterized by aging, wearing and other environmental dependencies. Abstract: Dependability assessment is a crucial activity to ensure the correct operation of complex systems. The output of dependability assessment activities include the quantification of reliability, availability, maintenance and safety related metrics. These metrics can assist in the identification of the system weak points or in the conception of mitigation strategies to increase the system dependability level. The development of advanced computer-aided methodologies to support dependability assessment activities is essential to automate and reduce the efforts implied by this process and similarly, the development of accurate dependability assessment methods is very important to increase the quality of the results. In this context, it is possible to identify different contributions that improve the dependability assessment through general-purpose modeling methodologies. However, existing solutions are ad-hoc applications specified with low-level stochastic formalisms and this complicates their adoption in the industry. Accordingly, this paper presents Stochastic Hybrid Fault Tree Automaton (SHyFTA) based simulation algorithm that allows the accurate dependability analysis of repairable multi-state systems. SHyFTA integrates the stochastic and deterministic operation of the system under study as well as their interactions. The algorithm is formalized through an object-oriented software architecture, which is developed as a software library for the modeling and simulation of repairable SHyFTA models. Following the proposed architecture, a Matlab® implementation of this library, SHyFTOO, has been developed and validated with a thorough test campaign. In order to provide a guideline to the end-users and show the potential of the SHyFTOO library, the case study of a feed-water pumping system is implemented in detail and it is used to evaluate different preventive maintenance policies. The SHyFTOO library can open the way to further investigations that address the interactions between the failure behavior and the functional operation of a system and their combined effect on system dependability.