Research On State Recognition Of Freeway Traffic Flow Based On Toll Collection Data

Massive highway toll data contains rich information of the traffic flow state and operation features,using data mining technology for mining large amounts of data to extract the traffic flow state and operation features of meaningful information is significant in traffic industry of big data technology.Without adding traffic detection equipment in the case,the massive charges data mining,extraction of traffic state parameters for traffic status identification to solve the occasional traffic congestion on the highway is important.In this paper,the travel time of the entrance and exit information is extracted from the toll data,and the travel time is divided into two parts: the ramp and the main line.Secondly,based on the uniform acceleration hypothesis designing experiment of ramp travel time estimation in highway to solve travel time of the entrance ramp,and putting forward the method according to the proportion distribution of the road sections based on the congestion point analysis to estimate the main line of travel time,and travel time reliability is discussed by using charge data.Then,by using the model of sub-time and sub-road segmentation,the author calculates the travel speed of the traffic state parameters,the two-dimensional matrix of interval flow and interval density.Finally,the method of drawing traffic flow diagram is designed,and the traffic flow diagram is constructed to provide support for traffic status identification in the end.The innovation of this paper is as follows:?1?Based on the uniform acceleration hypothesis designing ramp time experiment,the travel time of the entrances and exits is analyzed,and the feasibility of the model is verified by the experimental data.On this basis,the main line travel time estimation model according to the proportion distribution of the road sections based on the congestion point analysis is proposed.?2?The sub-time and sub-road segmentation model is proposed,and the driving trajectory of the single vehicle is analyzed.The two-dimensional matrix of the traffic flow parameter is constructed.Different from the classical density formula,in this paper,from the definition of interval density,this paper uses the method of sub-time and sub-road segmentation and statistical period t_i+n^*?t of [t_i,t_i+1] as the reference cycle time.Instantaneous moment,a single car on the road section of the road belonging to the situation,and then get all the vehicles at the moment of the road distribution.Finally,the density of the statistical cycle reference time is obtained by using the ratio of the total number of vehicles and the length of the road section in the reference section of the different statistical periods,which is a variable independent of the travel speed and the intermittent flow rate.?3?Using the model to obtain the two-dimensional matrix of traffic flow parameters,draw the flow-density scatter plot and the velocity-density scatter plot,according to the free flow and congested flow of Pearson coefficient and different flow crowded and congested flow variation coefficient difference,and designing the steps of constructing the traffic flow diagram based on the three-phase flow theory,and provide the basis for the traffic state identification.