| Reliability analysis has important guiding significance and application value for product design,manufacture and fault diagnosis.In various systems of rail vehicles,components are closely related to each other.The existing reliability calculation methods have certain limitations in describing the relationship between components,and there are also problems that require a large amount of calculations.Accurately describing the relationship between components and building a reasonable reliability calculation method can obtain more accurate reliability data,thereby evaluating the reliability of the entire system.Based on this,this paper studies the system reliability analysis method based on the GO method,from the GO algorithm considering the feedback signal,the simplified GO model,the mapping conversion of the GO model to the Bayesian network,and the reliability of the air conditioning system of the EMU.The main research contents are as follows:Aiming at the problem that the traditional GO method directly disconnects the feedback when dealing with GO models with feedback signals,which reduces the accuracy of the analysis results and causes deviations from the actual situation,a feedback signal operator method is proposed.First,give the feedback signal codes X1～Xn,establish the signal equivalence relationship in the closed loop formed by the connection of the feedback signals,and establish the equivalence equation by using the power law of Boolean algebra operation.Secondly,disconnect the feedback signal,and calculate the signal flow probability when there is a feedback signal along the signal flow of the GO diagram until the feedback signal is generated,and obtain a system of n simultaneous equations to obtain the value of the feedback variable.Finally,a new operator 5A is defined,and the combined value of 5A feedback value and corresponding input signal is given to establish the GO model with feedback signal,and this method is applied to the establishment of GO model in the cooling mode of the air conditioning system.When using the GO method for reliability modeling,the units with a very small failure rate can be omitted directly to simplify the structure and calculation of the GO model.In view of the subjectivity of determining the units to be omitted in the establishment of the GO model of the system in the past,a GO model is proposed that uses the AHP to calculate the weight of system unit failures in the overall system failures and omits units with very small weights.Simplified method.Through this method,the failure weight of each component in the cooling mode of the air conditioning system is analyzed,and the optimization of the GO model is realized.Aiming at the problem of building GO model to solve the problem of large amount of calculation for the reliability of complex engineering systems,a conversion algorithm from GO model to Bayesian network is proposed.Firstly,the mapping conversion rules from common operators of GO method to Bayesian network nodes are formulated,and the conversion process from GO graph to Bayesian network is given.Secondly,the GO graph of cooling mode of airconditioning system is converted by this method,and the Bayesian The reliability data of the system is obtained by reasoning and calculation of the network computing software,and the reverse reasoning of the Bayesian network posterior probability is used to determine the weak link of the system.The system reliability calculated by the fault tree method is compared with the consistency of the failure probability sequence of the weak components The test verifies the effectiveness of the method used in the reliability analysis of the air conditioning system.The analysis results show that the reliability analysis method based on GO method is reasonable and superior in system reliability modeling,reliability calculation and system fault diagnosis,and can provide technical reference for reliability modeling and analysis of similar engineering systems. |
