Visual Analysis Method Of Abnormal Travel Research On Public Transportation

In recent years,with the rapid development of urban public transportation systems,the public transit traffic is advancing by leaps and bounds.Meanwhile,the accompanying abnormal traffic situations in public transportation are increasing,which have greatly hindered the safety of public transportation,such as theft incidents on buses,advertising harassments on subway and begging inside the subway or in the waiting room,especially in hot areas during peak hours.Therefore,it is essential to ensure the safe operation of public transportation by effectively combating abnormal travel individuals and groups.Fortunately,the ticket checking system can accurately record the daily travel information of the passengers,which provides direct and effective help for detections of abnormal transit.Retrieving the target information with specific characteristics using the travel data stored in the ticket checking system,which has been widely used in the process of public security organs handling public traffic safety incidents and related cases.However,the huge and complicated data information has increases the difficulty of retrieval work of the staff.As a result,how to efficiently retrieve valid information from public transportation data and display the retrieved information in an interactive and visual way are important research issue facing the relevant departments.In this background,based on the big data of urban passengers traveling in public transportation,this thesis analyzes the general rules of public transportation passengers,extracts abnormal travel characteristics.In addition,for the demand of criminal investigation business,we propose a visualization method of spatio-temporal feature retrieval and a multi-view interactive exploration method.Finally,a visual analysis system is designed and implemented using modern visual interaction technology,which integrates visual retrieval and analysis of abnormal behaviors.The main content of the thesis can be summarized as the following three aspects:(1)In order to intuitively and quickly retrieve the abnormal individuals and groups in public transportation,a visualization method of spatio-temporal feature retrieval is proposed.The visual retrieval method retrieves the travel rules of the suspicious travel cards and the associations among abnormal individuals and groups according to mastered time,location and other characteristic clues.(2)In order to verify the abnormality of passenger travel from various aspects,further identify the identification accuracy of an abnormal card in(1),a visual analysis method with multi-view interactive exploration of specific card travel information is proposed.The visual analysis method can perform specific exploration on time,region and station for one or more suspicious cards,then provide comprehensive and detailed visual information of individual travel.(3)Facing the demand of criminal investigation business,a public transportation visualization system integrating the spatio-temporal retrieval and visual analysis is designed.The system improves the consistency and efficiency of abnormal identification work by performing visual analysis while retrieving,and helps users complete a series of tasks such as abnormal identification through flexible interaction.In summary,this thesis realizes the visual retrieval and analysis of abnormal travel in urban public transportation,and develops a complete visualization system.The research results can be applied to related analysis of public transportation,and the developed system can be directly applied to actual Beijing card data,which provide scientific basis for abnormal travel detection,warning and visual control.