Estimation Traffic State For Highway Based On Floating Car Data

With the rapid development of China's regional economy,the scale of the expressway network has been expanded and completed,which makes expressway the main mode of transportation for short-and long-distance passenger and freight transport.However,with the increase rate of expressway utilization,the capacity and service level of expressway are obviously decreased,which also directly leads to the increase of traffic delay,the increase of travel time,the high occurrence of traffic accidents,the environmental pollution and so on.It is the most important to make full use of the advanced intelligent traffic system to achieve the intelligent traffic management and control of the expressway.Real-time accurate traffic state estimations is the key to expressway intelligent management and control,and is also the basis for intelligent traffic control,analysis,event detection and accident warning of expressway,which is of great significance to the construction of expressway ITS.Probe vehicles are one of the most effective way to collect road traffic data with the characteristic of their wide coverage and easy collection.Probe vehicles equipped with global positioning system(GPS)can collect information such as vehicle position,instantaneous velocity.By establishing the method of probe vehicle-based traffic state estimation of expressway,the traffic flow,density and speed of expressway are estimated accurately in real time,so as to provide certain theoretical basis and technical support for formulating efficient management and control strategy and realizing the user information service of expressway.Firstly,the overall architecture of expressway traffic state estimation model based on floating car data is proposed.Compared with other expressway traffic state estimation models,the model could make full use of the traditional traffic detector data and floating car data,and the calculation method is simple and easy to implement,so that it is more accurate to obtain the real-time traffic state information of expressway.Secondly,on the basis of analyzing the expressway empirical data,K-means clustering algorithm is used to address the problem of sample selection bias.Then,apply least square method based on function forms to realize the calibration results of the traffic speed-density relationship,which is devoted to traffic state estimation of expressway.Thirdly,introduce the core modules of probe vehicle-based traffic state estimation of expressway in detail,including the cumulative vehicles estimation module and traffic state estimation module.According to the information of basic road edge and floating vehicles,a nonlinear programming model is established.The minimum of traffic speed estimated error is taken into consideration to build an objective function,the vehicles and road edge topology structures are taken as constrains,and then cumulative vehicle estimation is put forward.Afterwards,state estimation model of expressway is raised according to the definition of traffic parameters.This is also the main core of this paper.Finally,the simulation experimental scenario is built by applying the software of SUMO,and the effect of the constructed traffic state estimation base on probe vehicles model is evaluated,mainly including the error of traffic flow parameters and the evaluation analysis of the constructed traffic state estimation model under different probe vehicle penetration rate,which provides an effective way to realize the traffic flow state estimation of expressway.