The Study And Realization Of Railway Tunnel Security Management System Based On Data Mining

The operation of railway tunnel Security is an extremely complicated nonlinear relation that the running of tunnel makes an influence and interconnection with the environment,as well as an important link to guarantee the normal operation of railway, which plays an important significance to make certain of absolute Security of railway transportation. This article focuses on the collation and maintenance of accumulated and abundant date, and the processing of association rule which is fully excavated from crucial data by adequately making use of computers in order to find something available and significant knowledge about management and decision, resulting in better maintenance of the tunnel safety, effective monitoring of disease, immensity reduction of maintenance costs and the most possible cultivation and utilization of present knowledge guaranteeing the Security of railway transportation.This article, first of all, briefly reviews the relational theory of data mining focusing on the introduction of Apriori algorithm which is widely used in the association rule technology.It presents the main ideas, the steps and the pseudo code of the algorithm which makes the algorithm realization. In addition, the basic terminology and category corresponding to the project are introduced as well. The paper, what's more, focuses on the construction of management system of tunnel lining. Therefore, in the article,it takes the basic steps strictly according to the information system project that put forward by the Ministry of industry to system development, namely the typical requirements analysis stage, collectivity design stage (a summary of the design phase), detailed design stage of the system and implementation stage of the system. In the requirements analysis stage, on the one hand, the paper emphasize on the actual requirement of the project and its transformation into the need of the information system. On the other hand, the paper research into the feasibility analysis, performance requirements analysis and pay more attention to the demand analysis of the function of the system, including the system login, user management and permissions settings, furthermore, the demand analysis is also in possession of the basic information management of tunnel, radar information management, diseases information management (the definition of management here contains the function of append, delete, modify, query, specially the function of supporting data import in batches), graphical statistical analysis of tunnel deterioration information, association rules mining of the data, security evaluation of tunnel and the function of decision. In the design stage of the system including a summary of the design phase and the detailed design stage, the article attaches importance to the latter stage that the sequence diagram and activity diagram of each of function as well as the flow of disposal of the system are drawn and so on according to every requirements analysis making a solid foundation for achievement of the next phase of the system. In the implementation stage of the system, which is the actual implementation of the system, Studio platform technology of Microsoft and SQL Server data are adopt to realization the technical which is analysis and design above, and almost meeting the demand hereinbefore. In the tunnel safety assessment and decision-making function, the system is carrie d out to achieve stratified analysis.First,the system dinstinguishes different linging,spec ify different types of evaluation and support the evaluation of multi-section,integrating t he results of the score of each section to obtain overall conclusions of the safety of the t unnel.In decision-making function,it firstly designs the assessment of weight for decisio n-making,entering the value of the upper triangular matrix system to automatically deter mine the consistency, logic, operations to come to the proportion of the weight of each r ow. The second step is the set of criteria layer weights,caculating according to a two-tier value weight and the maintenance decision of recommendations and conclusions propo unded by the system.To perfect Apriori algorithm to better fit the systematic classification model mentioned above, this paper, based on a proved algorithm theorem, makes several rewarding attempts on branch-cutting and I/O spending reduction and gives a satisfying data mining result which is demonstrated in this project.