Characteristic Analysis Of Urban Highway Traffic Accidents Based On Data Mining

With the rapid development of China’s transportation,The rapid growth of traffic accidents has become a serious problem facing traffic management.In all kinds of traffic accidents,urban roads are the most densely-existing places of modern traffic,the occurrence of urban road traffic accidents will lead to serious negative effects,the safety of the people and society will also face a great threat.Prevention and reduction of urban road traffic accidents has always been the focus of the relevant management departments,and is also an important research topic of domestic and foreign scholars.The historical data of traffic accident implicitly contains the relationship of various influencing factors in the process of traffic accident.By using data mining to study it,the discovered knowledge can be used as the theoretical basis of urban road traffic safety management and traffic safety education,and provide scientific support for relevant policies.The core idea of data mining technology is to find out the hidden law among the massive data.At present,there are a lot of original data of traffic accidents in various traffic management departments.These data can be studied by data mining,so the mining and analysis of urban road traffic accident data has certain theoretical significance and social value.This thesis first clarifies the relevant concepts and standards,including defining the connotation of urban road traffic accidents,discussing the feasibility of data mining technology applied to the analysis of urban road traffic accident characteristics,and defining the selection criteria of data.On this basis,the relevant data was obtained through investigation,and an urban road traffic accident data set containing 13 fields was constructed to conduct data mining research on the influencing factors of urban road traffic accidents.Using descriptive statistics to analyze the distribution characteristics of urban highway traffic accidents under a single influencing factor are analyzed,and it is found that urban highway traffic accidents show obvious time characteristics,vehicle type characteristics and weather characteristics;using field selection to construct different data set according weather,road segments and vehicle types,and then,the Apriori algorithm is used to mine the overall association rules of urban road traffic accident influencing factors and conduct a comparative study on the accident association rules of urban highway traffic accidents in different situations.It helps a lot to understand the internal mechanism of urban highway traffic accidentsThe research results of this thesis can guide relevant management departments to carry out urban highway traffic management.At the same time,the analysis of other traffic accidents can use the methods in this thesis as a reference.