| With the building of Smart City,the traffic administrative department has accumulated a large number of traffic accident data.It is of great significance to mine the traffic accident patterns and laws contained in these data for reducing traffic accidents and improving transportation safety.However,it’s very challenging to analyze traffic accident data because they are inherently complex.Visual analytics,which combines human cognitive ability with machine computing capability through visual interactive interface,is an important tool for analyzing complex data.Therefore,this dissertation studies the visual analytics method of traffic accident data in conjunction with the research objective of the program named "Analysis and Management System Based on Traffic Accident Data",which is funded by the Anhui Provincial Department of Science and Technology.The main research contents are as follows:(1)Research on visual analytics method of urban traffic accident spatial-temporal pattern: A visual analytics system for spatial-temporal pattern of traffic accident is designed and implemented.It’s combined with urban road network data and historical weather data,and uses several linkage views to show traffic accident comprehensively in time and space.In particular,a new recognition method based on road accident risk is proposed for black-spot which is an important spatial pattern of traffic accident.It can assist users to explore black-spot in different time or weather by cooperating with other views in the system,so as to provide targeted suggestions and assistance for the implementation of traffic safety.(2)Research on the enhancement and application for the semantic feature of traffic accident space: The in-depth analysis of the traffic accident spatial pattern has been affected because of the lack of the semantic description for the accident location in traffic accident data.Therefore,this dissertation proposes a method for spatial semantic enhancement and analysis of traffic accident data.In the method,the spatial semantic information of traffic accident points is enhanced by POI data and is analysed with a method based on Self-Organizing Map(SOM)algorithm.Based on the analysis results,a visual analytics system is designed and implemented.Through specific color coding,view linkage and interactive analysis,the system helps users reveal the spatial semantic patterns of traffic accident,so as to deeply understand the causes of traffic accident. |
PDF Full Text Request