Continuous-Time T-S Dynamic Fault Tree Analysis Method And Its Applications

Fault tree analysis is a common method for reliability analysis,fault analysis and diagnosis.Discrete-time Dugan dynamic fault tree analysis method solves the problem that the static fault tree cannot describe the dynamic characteristics such as time series,redundancy and correlation.However,the discrete-time Dugan dynamic fault tree analysis method has analytical errors and cannot be directly quantitatively analyzed.Discrete-time T-S dynamic fault tree analysis method further improved the ability of the fault tree to describe static and dynamic failure behaviors,and can be directly quantitatively analyzed,but there are still analytical errors and cannot reflect the real-time change trend of system reliability.To this end,continuous-time Dugan dynamic fault tree analysis method is proposed,and continuous-time T-S dynamic fault tree analysis method is proposed.Based on this,continuous-time T-S dynamic fault tree importance analysis method is proposed.Firstly,in order to solve the disadvantages that the discrete-time Dugan dynamic fault tree analysis method has analytical errors and cannot be directly quantitatively analyzed,Continuous-time Dugan dynamic fault tree analysis method is proposed.Dugan dynamic gate and unit step function are used to describe the relationship between events and their timing fault relationships.Combined with the integral characteristics of impulse function,continuous-time Dugan dynamic fault tree analysis method based on superior event failure probability density function algorithm is presented.The feasibility of the proposed method is verified by comparing with discrete-time Bayesian network solving the Dugan dynamic fault tree.Secondly,in order to solve the disadvantages that the discrete-time T-S dynamic fault tree analysis method has analytical errors and can not reflect the real-time change trend of system reliability,continuous-time T-S dynamic fault tree analysis method is proposed.Firstly,continuous-time T-S dynamic gate describing the static and dynamic failure behaviors is defined and construction method of continuous-time T-S dynamic gate is proposed,and then a method for solving the rule execution probability and the upper event failure probability function is proposed.The feasibility of continuous-time T-S dynamic fault tree analysis method is verified by comparing with discrete time T-S dynamic fault tree analysis method and the Markov chain solving Dugan dynamic fault tree.Then,in order to provide reference for the weak links discovery and reliability improvement,on the basis of continuous-time T-S dynamic fault tree,continuous time T-S dynamic fault tree importance analysis method is proposed,including probability importance,criticality importance and improvement function.The importance of components in the system is reflected from the change of fault state,the rate of change of fault probability and the improvement of system reliability.The feasibility of the proposed importance analysis method is verified by comparing with the traditional fault tree importance analysis method and discrete-time T-S dynamic fault tree importance analysis method.Finally,the proposed continuous-time T-S dynamic fault tree and importance analysis method are applied to the reliability analysis of the wheel brake system,and the failure probability and probability importance,key importance,improvement function of each component within a given task time are obtained,which provides reference for the weak links discovery and reliability improvement.