Traffic State Identification And Association Analysis Of Urban Area Based On Trajectory Data

In the management of public security traffic,it is very important to accurately and quickly identify the traffic state of the road network.This enables managers to keep abreast of the state of the whole city and deal with emergencies in time.For this reason,the traffic management department has built a large number of information acquisition equipment,which is expensive,difficult to maintain and easy to have adverse effects on the environment.At the same time,the traffic management department has accumulated a large amount of urban road network traffic state data,but these data can not be fully utilized.The emergence of Internet-based vehicle offers a new choice for urban traffic state identification.The features of long running time,wide coverage and no need to install special equipment provide an economical and fast method for urban traffic state identification.In this paper,we use the trajectory data to identify the traffic state of urban areas,and analyze the association of the obtained regional traffic state data to mine the hidden rules of traffic state data.Firstly,this paper reviews and summarizes the current domestic and foreign research on traffic state identification and association analysis,clarifies the research content and research methods of the paper.Secondly,the paper introduces the content of the trajectory data of Internetbased vehicle and preprocesses the data.Then,using programming interface ArcPy in ArcGIS platform automatically realizes the map matching of massive trajectory data.Finally,the papr obtains the combination of trajectory data and road segments.Secondly,considering the influence of sample size on calculating average travel speed in fixed time intervals,a method of calculating average travel speed based on sample size is proposed.This paper analyses the characteristics of the existing traffic state identification methods,and combines with the target of traffic state identification in urban areas.Referring to the method of GB/T 33171-2016 Standard for Evaluation of Urban Traffic Operational Conditions,the traffic state identification is carried out by using the average travel speed and free flow speed of road segment.In the case analysis,for the matching results of more than 111 million maps,the proposed method is used to identify the traffic state,and the obtained road network traffic state data is visually analyzed.Finally,this paper proposes an improved data preprocessing method,which transforms traffic state data into temporal,spatio and spatio-temporal transaction data sets.Four constraints rules are proposed according to the actual traffic operation characteristics.FP-Growth algorithm is used to mine the temporal,spatio and spatio-temporal association rules in accordance with the traffic operation characteristics by combining the corresponding constraints rules.The example verifies that the improved FP-Growth algorithm has at least 86% improvement in mining efficiency compared with Apriori and FP-Growth,which only mine frequent binomial sets.The mining association rules are more analytical and prove the effectiveness of the improved algorithm.