Reliability Modeling And Evaluation On The Synthetic Automation System Of Marshalling Yard

Under the new situation that railway is transforming into modern logistics company and freight volume has increased significantly,the synthetic automation system of marshalling yard came into being.However,the related reliability research work is relatively lagging,the system improvement and application guidance lack of the theoretical support.Based on the review of the theory and practice of reliability evaluation of complex large-scale system,the research and development of the synthetic automation system of marshalling yard and its reliability research status,the paper takes the synthetic automation system of marshalling yard and its users as the research objects,and carries out the modeling and evaluation of system reliability and human factor reliability.In view of the shortcomings of traditional reliability analysis methods in the analysis of complex large-scale systems,such as high computational complexity,dynamic failure correlation,lack of failure data,etc.,the paper proposes a reliability evaluation method for the synthetic automation system of marshalling yard based on Dynamic Bayesian Network,which can also be applied to the reliability evaluation of other complex large-scale systems.First,based on the analysis of system structure and boundary,the dynamic fault tree model of the synthetic automation system of marshalling yard is established,which is the abstraction of system faults;the probability distribution of bottom event is determined by the probability analysis method based on fuzzy set theory,which solves the problem of lack reliability data and test data,and provides support for follow-up quantitative analysis.Secondly,the dynamic fault tree of the system is transformed into the Dynamic Bayesian Network model,which has the more stronger reasoning and expression ability,and the factors such as coverage factor,common cause failure,redundant structure and maintainability are considered to modify this model.Finally,the Dynamic Bayesian Network direct inference algorithm is used to carry out two-way reasoning on the model,and it is concluded that the reliability index of the synthetic automation system of marshalling yard is not lower than that of the similar signal products,and its reliability fully meets the site requirements.At the same time,it is necessary to improve the dual computer switchover equipment and control common cause failure.In view of the shortcomings of traditional railway human reliability analysis methods,such as not considering the cognitive behavior process and not considering the behavior forming factors sufficiently,the paper proposes a human reliability analysis method based on cognitive behavior process,which can also be applied to the human reliability analysis of other two-person mutual control modes.First,based on the analysis of human cognitive behavior process,the behavior process of operators is divided into four steps: awareness,diagnosis,decision-making and implementation,and the fault recovery mode under the two-person mutual control mechanism is given.Secondly,the paper gives a quantitative calculation method for different types of human error classified by SRK model under different behavioral processes.Finally,by comparing the probability of human error under the same situation before and after the implementation of the synthetic automation system of marshalling yard,it is concluded that the improvement of information quality and the control of the implementation process can effectively improve the human reliability.This paper fills the gap of reliability evaluation of complex large-scale system and human reliability evaluation under the mechanism of two-person mutual control in railway field.By using the suggestions put forward in the paper to improve the system reliability and human reliability,we can take measures to further the system structure and configuration,improve the interface content and safety control function,which has practical guiding significance for the in-depth development and upgrading of the system.It also provides effective technical support for system certification in the future.