Research On Dynamic Fault Tree Quantitative Reliability Analysis Method Based On Deterministic And Stochastic Petri Net And Monte Carlo Simulation

ABSTRACT:The rapid development of computer science helps increase the system integration density and the degree of automation in the field of aerospace engineering, traffic engineering, electric engineering and medical facilities. Meanwhile, the system scale and complexity has also been augmented. Since these systems requirement for safety and reliability is so strict, relative safety assessment and reliability verification has drawn lots of attention. Specifically, modeling analysis of complicated system with dynamic process has drawn the attention of designers and operators. Though researchers have did many efforts in terms of standards and plans, deficiency of the applicability and accuracy for complex systems still exits. Especially for the issues such as method applicability for event randomness, model scale control, algorithm complexity and feasibility of automation, there is space for improvement.According to the statements above, the purpose of this thesis is to come up with a quantitative dynamic fault tree analysis method based on Deterministic and Stochastic Petri Net and Monte Carlo simulation. This method, based on the Deterministic and Stochastic Petri Net, through the statistics generated by system fault tree, combines three simulation platforms together. Besides, new dynamic fault tree models, date structures, model solutions, simulation parameter estimation methods and reliability index calculation are created and used to realize the huge and complex dynamic fault tree quantitative analysis. Innovation specifics are as below:Firstly, the Petri Net model in this thesis has taken into consideration different dynamic logic gate representing system complicate dynamic characters including sequence, independency, repairs and spares. Further discussion is made in terms of model applicability and correctness. Besides, proper Petri Net simulation platform is chosen for model quantitative calculation.Secondly, in the part of fault tree Monte Carlo simulation, new fault tree data structure, structure function generation method and calculation method are created. For different kinds of basic events, corresponding sampling method and how to decide the sampling frequency are both created. What’s more, related reliability index calculation method is proposed according to simulation statistics.Last but not least, by chosen typical fault tree examples and thorough complexity analysis, the new quantitative method is proved to be more accurate and faster than the analytical method and its extended method. Further, the new method is applied to the analysis of huge fault tree of CBTC Zone Controller MA calculation module. The system timed failure CDF and event importance are calculated. Then, the result effects upon the system design and operation maintenance are carefully examined.