Research On Association Rules And Evolution Behavior Of Railway Accident Causes Based On Complex Network

The ever-increasing operational mileage of railway and the demand of passenger and freight transportation make people put forward higher requirements for the safety of railway transportation.Under the background of the complexity and nonlinearity of the railway system,the mechanism of railway accidents is more complicated.Mining and learning effective information from historical accident data are of great significance to improve the system security.However,the current railway accident data is mostly stored in the form of text.It is difficult for the researchers and managers to identify and use the potential information in a large amount of accident text data effectively and fully.In order to process a large amount of accident text data in railway complex system efficiently and accurately,the analysis method of association rules and dynamic evolution behavior of railway accident causes is put forward using text mining,association rules mining and complex network analysis.On this basis,focusing on the personnel operation process in railway transportation production activities,the relationship between causes and the law of railway accidents are studied in depth taking the railway operation accident as an example,which can provide references for the decision-making of the railway transportation safety management.The main research contents are given as follows:(1)Using the text mining method,the mining process of railway accident causes is put forward.The definition and classification of railway operation accidents are proposed.Taking a total of 2,618 railway operation accident investigation reports from2016 to 2018 as an example,71 railway operation accident causes are identified.The causes of each accident are mined based on Python programming and a structured data set is established,which provides data support for the subsequent research.(2)A process of mining association rules of railway accident causes based on Apriori algorithm is designed.In view of the insufficiency of traditional association rules analysis methods,a method of analyzing the association rules of railway accident causes based on complex network is proposed.Taking the above-mentioned data set of railway operation accident causes as an example,the Apriori algorithm is used to mine the strong association rules between causes,and the high support association rules and the high confidence association rules are analyzed.On this basis,a railway operation accident causes correlation network consisting of 118 nodes and 334 edges is constructed based on the complex network theory.Through the view analysis,community division and topology analysis of the network,the key causes of railway operation accidents and the internal relations between the causes are discovered.The results show that the association rules exhibit obvious aggregation characteristics,and the proposed analysis method of association rules further reveals the hidden information in the mining results.(3)A modeling and simulation method for the dynamic evolution behavior of railway accident causes is proposed based on complex network dynamics and the improved infectious disease model.Considering the importance of nodes,the improved infection level parameters and recovery level parameters are proposed,and the states and evolution process of the elements in the system are defined.Then,a phased dynamic model is constructed and the corresponding simulation method is designed.Based on the data set of railway operation accident causes,a railway operation accident causes evolution network is constructed taking the confidence of association rules as the weight of edges.By simulating evolution behavior of the causes under different parameters and different network structures using Python programming,the validity of the model is verified,and the difficulties in the prevention of railway operation accidents are revealed.