Estimation Of Urban Traffic State Based CTM-v And Vehicle Trajectory Data

With the increasing perfection of urban road infrastructure and the increasing number of vehicles on the road,traffic congestion has become a major problem affecting People’s Daily life.However,real-time and accurate state parameters are the basis of traffic management and control strategies.Therefore,Urban traffic networks state estimation is of great significance to alleviate traffic pressure and improve traffic conditions.This paper takes the urban traffic networks as the application background.and uses the vehicle trajectory data.This paper takes uses the vehicle trajectory data,such as speed,to estimate the traffic state,and the method of velocity estimation is studied.In order to simplify the estimation process,the traffic flow model of urban road network is established,which is helpful to analyze the temporal and spatial evolution law of road speed and provide data support for section traffic status discrimination,traffic management and control.The main content of this paper includes the following parts:Firstly,the cell transmission model with velocity as the state is established,which is called network cell transmission model for velocity(CTM-v).This model can directly apply the velocity data of sampled vehicles to build the system output equation,so it is easier to build the system state estimation problem accurately.Secondly,in view of the problem that estimation of traffic state based on traffic model is easy to deviate from the real value,while estimation of traffic state based on vehicle trajectory data has low permeability in some sections and is interfered by noise,a filtering method based on traffic flow model and real-time data is proposed to estimate traffic state.The filtering estimation method combines the measured noise data into a traffic flow model to obtain real-time traffic state estimation.Considering that the cellular model of urban road network speed is a typical nonlinear system and the model is non-smooth,the Ensemble Kalman Filtering method is used to further estimate the velocity state.Furthermore,because the use of vehicle trajectory data may cause serious privacy problems,this paper proposes how to obtain the cellular travel time and speed data of the modeled road section from vehicle trajectory data and protect the privacy of the data.Firstly,the vehicle trajectory data used in this paper is introduced.Secondly,a data space acquisition method suitable for traffic modeling is proposed to show that the data are sampled in those location areas.Finally,the privacy protection method of vehicle trajectory data is introduced to meet the data requirements of privacy protection and fine-grained urban traffic modeling applications using GPS data.Finally,aiming at a road network containing 17 intersections,taking road network control and traffic flow parameters as inputs,MATLAB is applied to establish a road network filtering algorithm to analyze the state evolution of the road network and verify the Ensemble Kalman filtering algorithm.Comparing the influence of GPS data sampling method on state estimation results.