1,721,036 research outputs found
The Conversion of Dynamic Fault Trees to Stochastic Petri Nets, as a case of Graph Transformation
No full textAbstractA model-to-model transformation from Dynamic Fault Trees to Stochastic Petri Nets, by means of graph transformation rules, is presented in this paper. Dynamic Fault Trees (DFT) are used for the reliability analysis of complex and large systems and represent by means of gates, how combinations or sequences of component failure events, lead to the failure of the system. DFTs need the state space solution which can be obtained by converting a DFT to a Stochastic Petri Net: this task is expressed by means of graph transformation rules, and is applied to a case of system
Development of a Dynamic Fault Tree solver based on Colored Petri Nets and graphically interfaced with DrawNET
Full text linkSAN models of a benchmark on dynamic reliability
Full text linkThis report provides the detailed description of the Stochastic Activity Network (SAN) models appearing in [1] and concerning a benchmark on dynamic reliability taken from the literature
Extended Fault Trees Analysis supported by Stochastic Petri Nets
Full text linkThis work presents several extensions to the Fault Tree [90] formalism used to build models oriented to the Dependability [103] analysis of systems. In this way, we increment the modelling capacity of Fault Trees which turn from simple combinatorial models to an high level language to represent more complicated aspects of the behaviour and of the failure mode of systems. Together with the extensions to the Fault Tree formalism, this work proposes solution methods for extended Fault
Trees in order to cope with the new modelling facilities. These methods are mainly based on the use of Stochastic Petri Nets. Some of the formalisms described in this work are already present in the literature;
for them we propose alternative solution methods with respect to the existing ones. Other formalisms are instead part of the original contribution of this work
Generalized Fault Trees: from reliability to security
Full text linkFault Trees (FT) are widespread models in the reliability field, but they lack of modelling power. So, in the literature, several extensions have been proposed and introduced specific new modelling primitives. Attack Trees (AT) have gained acceptance in the field of security. They follow the same notation of standard FT,but they represent the combinations of actions necessary for the success of an attack to a computing system. In this paper, we extend the AT formalism by exploiting the new primitives introduced in specific FT extensions. This leads to more accurate models. The approach is applied to a case study: the AT is exploited to represent the attack mode and compute specific quantitative measures about the system security
The conversion of Parametric Dynamic Fault Trees to Stochastic Well-formed nets as a case of graph transformation
No full textThe Fault Tree (FT) is a stochastic model for the reliability analysis of complex and large system: it allows to model as a Direct Acyclic Graph (DAG) whose structure is similar to a tree, how combinations of component failure events determine the failure of subsystems or of the whole system; FT is a bipartite graph: its nodes can be event nodes or gates; event nodes model the occurrence of failure events, while gates propagate the failure towards the upper level event nodes if a particular logic condition is verified; the lowest level event nodes represent the failure of the basic components of the system, the internal event nodes represent the failure of subsystems, while the root node models the failure of the whole system. FT can be easily analysed in a combinatorial way, so FT is a widespread model for the reliability analysis, but it suffers from some modeling limitations, such as the assumption of statistical indipendence among events; for this reason, an evolution of such model, called Dynamic FT (DFT) has been proposed with the aim of modeling event dependencies and more complicated failure propagation modes. The combinatorial analysis is not enough for DFT: it needs also the state space analysis; the generation of the state space of a DFT may be complicated, while the state space of a Stochastic Petri Net (SPN) can be generated in a direct way; so, an approach for the DFT analysis consists of translating a DFT in a SPN modeling the same failure propagation mode. In the case of Parametric DFT (PDFT), i. e. DFT where a unique parameterised subtree represents compactly the failure propagation mode of several identical subsystems, the translation result is a Coloured SPN in the form of Stochastic Well-formed Net (SWN). The translation of PDFT in SWN can be performed by means of transformation rules: PDFT nodes can be events or gates, so for every kind of event or gate, a rule for its transformation to SWN is defined; the starting graph is a PDFT and at each step of the transformation, an event or a gate is replaced by the corresponding SWN applying the relative rule; when the transformation process ends, we obtain a SWN that models the failure propagation mode of the whole starting PDFT. Now, the state space analysis of the PDFT can be performed through the corresponding SWN
BDD based analysis of Parametric Fault Trees
No full textSeveral extensions of the Fault Tree (FT) [1] formalism have been proposed in the literature. One of them is called Parametric Fault Tree (PFT) [2] and is oriented to the modeling of redundant systems, and provides a compact form to model the redundant parts of the system. Using PFTs instead of FTs to model systems with replicated parts, the
model design is simplified since the analyst can fold subtrees with the same structure in a single parametric subtree, reducing the number of elements in the model. The method based on Binary Decision Diagrams (BDD) [3, 4, 5] for the quantitative analysis of FTs, is adapted in this
paper to cope with the parametric form of PFTs: an extension of BDDs called Parametric BDD (pBDD) is used to analyze PFTs. The solution process is simplified by using pBDDs: comparing the pBDD obtained from a PFT, with the ordinary BDD obtained from the unfolded FT, we can observe a reduction of the number of nodes inside the pBDD. Such
reduction is proportional to the level of redundancy inside the PFT and leads to a consequent reduction of the number of steps necessary to perform the analysis. Concerning the qualitative analysis, we can observe that several Minimal Cut Sets (MCS) obtained from the FT model of a redundant system, involve basic events relative to similar components. A Parametric MCS (pMCS) allows to group such MCSs in an equivalence class, and consequently, to evidence only the failure pattern, regardless the identity of replicated components. A method to derive pMCSs from a PFT is provided in the paper
Applying Generalized Continuous Time Bayesian Networks to a reliability case study
Full text linkWe discuss the main features of Generalized Continuous Time Bayesian Networks (GCTBN) as a reliability formalism: we resort to a specific case study taken from the literature, and we discuss modeling choices, analysis results and advantages with respect to other formalisms. From the modeling point of view, GTCBN can represent dependencies involving system components, together with the possibility of a continuous time evaluation of the model. From the analysis point of view, any task ascribable to a posterior probability computation can be implemented, such as the computation of system unreliability, importance (sensitivity) indices, system state prediction and diagnosis
- …
