Research On Inventory Forecast And Dispatching Balance Of The Recycling Class Ticket In Urban Rail Transit Based On Data Mining

Under the background of network operation,the urban rail transit lines are gradually connected into a network,which brings great challenges to the operation and management.As an important part of the operation and management of urban rail transit,the management of recycling class ticket is a key factor to measure the service quality of rail transit and an important means to control the operation cost reasonably.Therefore,it is of great significance for the operating company to manage the recycling of ticket scientifically and proactively,in order to maintain normal operation and reduce operating costs.In this thesis,the recycling class ticket as the research object,a variety of data mining techniques are integrated to establish the inventory forecast model of ticket,then to optimize the dispatching problems of the recycling class ticket in urban rail transit.It is mainly for the following aspects:First of all,through the statistical analysis of the flow data,the sales data and the use proportion data,the thesis summarizes the basic features presented by the ticket data.The recycling process of the recycling class ticket are summarized to point out the existing problems and the future direction of improvement.Second,based on a variety of data mining techniques,the inventory forecast model is established.Combining the time series modeling method with the support vector regression model theory,a short-term forecasting model of the recycling class ticket is established.Using fuzzy information granulation method to process the data,the thesis establishes two combination models of ticket variance space forecast and discusses the application of the model,which can be used to grasp the change trend of ticket macroscopically and verify the result of short-term forecasting model.Using ARIMA-SVR short-term forecast model,the ticket sales and flow data are predicted,and the inventory safety range of the station is deduced through the inventory forecast model,which reduces the inventory cost effectively.Finally,the thesis studies the dispatching balance problems of ticket between station and station in urban rail transit,draws out the amount of ticket allocation according to the inventory forecast model,and proposes cycle calculation method on the goal of balancing the number of stations involved in the dispatching.Based on the feasibility of the metro as a freight mode,the thesis proposes that ticket can be dispatched by using the metro.The generalized cost is introduced to calculate unit freight rates and determine the route of dispatching.The dispatching model is established based on the transportation model and the dispatching plan is provided by using the actual operation data of Beijing Subway.