| With the progress of science and technology,5 G,artificial intelligence,big data,such as the Internet of things for a new generation of information technology and the social public life closer together,prompted the social public higher requirements of the city information,intelligence,wisdom arises at traffic,intelligent transportation,for the increasing urban congestion brings hope to resolve the problem.The real-time,rapid,accurate and stable intelligent identification of the spatial and temporal distribution characteristics of urban traffic status and the excavation of its regularity in time and space are the basis for the intelligent traffic to play its role.Based on the analysis of GPS data of taxi,this paper proposes a set of spatiotemporal distribution characteristics description method of city-level traffic state.The main contents of this paper are as follows :(1)through the analysis of the influencing factors of traffic state,it is found that the traffic state is greatly affected by the social activities of the public.According to the characteristics of public social activities,it is proposed that the change of traffic state has two characteristics: time-varying and periodicity;(2)Based on the traffic flow theory,the characteristics of taxis and road attributes,the traffic index that can reflect the traffic state of the section and the proportion of road space as well as the taxi number index is proposed.(3)Based on the similarity between road traffic network and Internet network,Page Rank algorithm is used to analyze the importance of road sections and construct regional traffic index that can reflect regional traffic status;(4)Based on k-means algorithm and the relationship between traffic index and taxi number index,complete the threshold search of traffic state;(5)The periodicity of traffic state is divided into daily periodicity within a week and week-date periodicity between different weeks.The periodic analysis method based on outliers is proposed,and the two kinds of periodicity are combined to excavate the regularly congested roads.Finally,the temporal and spatial distribution description system of traffic state is proposed based on the time variation of traffic state.The full sample GPS data of Xi ’an taxi in 2017 was used for analysis.(1)Time-varying analysis based on traffic index found that 24 h a day can be divided into 10 periods according to the change of traffic index and time-varying trend of traffic index,and the 10 periods can be divided into 5 stages according to the severity of traffic state.(2)based on periodic analysis found that Monday morning peak,peak on Friday night,and the other week Friday night time date at the same time period of more congestion,Monday morning peak,peak on Friday night the other week date concentrates on new gridlocked expressway and high-speed connection road,new traffic jam focused on Friday night in qujiang northern scenic area,the west side of the clock tower commercial circle and business circle.The main congested areas of Xi ’an are Zhangba North Road,Xiaozhai Business district and Bell Tower business district.(3)Based on the analysis of different types of roads,it is found that the traffic condition of xi ’an expressway is better than that of the main road and secondary road.In the second week of April 2017,the space of congested sections in the main road in the morning and evening peak hours reaches 17.42% and 20.13%. |
PDF Full Text Request