Traffic Flow State Estimation For Mixed Traffic Flow Of Connected And Automated Vehicles Based On Particle Filter

In recent years,the large-scale growth of highway construction mileage has promoted the rapid development of the national economy.While facilitating the travel of residents,it has also led to the frequent occurrence of highway traffic congestion and traffic accidents,which has caused huge losses to the social economy.Traffic flow state prediction on highways can reduce or avoid the occurrence of these phenomena to a certain extent.At the same time,with the increasing popularity of intelligent networked vehicles,before ordinary vehicles are completely eliminated from the market,a mixed traffic flow environment composed of intelligent networked vehicles and ordinary vehicles will exist for a long time.It’s unclear what effect the addition will have on traffic flow.This paper takes the mixed traffic flow environment where the intelligent networked vehicles and ordinary human-driven vehicles coexist as the research object,and uses the particle filter algorithm to estimate the macroscopic traffic flow parameters of the highway section,so as to realize the estimation of the highway traffic flow state.First,according to the following characteristics of intelligent networked vehicles and ordinary human-driven vehicles,the intelligent driver model is selected as the vehicle following model of ordinary vehicles,and the CACC model of PATH laboratory is selected as the vehicle following model of connected vehicles.Degradation phenomenon,introduce the market penetration rate of intelligent networked vehicles,build a mixed traffic flow model based on the equal relationship of constant speed and equal spacing when the traffic flow reaches an equilibrium state,and analyze the basic diagram;Secondly,the stability of the mixed traffic flow model in Chapter 1 is explored,and the stability performance of the mixed traffic flow model constructed in different speed ranges is analyzed and studied,which provides a theoretical basis for the micro-simulation experiments,and provides a theoretical basis for the penetration rate of different connected vehicles.,The stability of the combination of different speed and speed intervals is studied,and the proportion of the number of connected vehicles when the stability is the best is determined.Finally,based on the second-order macroscopic traffic flow model,combined with the mixed traffic flow model,a state-space model of particle filter traffic flow state estimation is constructed,and the state estimation effect of particle filter on mixed traffic flow is measured by Matlab and SUMO simulation software.The effect is evaluated and discussed,mainly including the tracking ability of the particle filter state estimation model to the fluctuation of the traffic flow state and the error comparison with the extended Kalman filter.