Microscopic Traffic Flow Modling Considering Heterogeneous Driving Behaviors And Traffic Impact Analysis

With the rapid growth of industry and economy,the number of motor vehicles has also increased in recent years.The traffic environment and safety problems have become sharply prominent,seriously affecting the quality of residents’ living.Car-following and lane-changing behavior as the basic vehicle’s behavior at microscopic level are the main factors affecting traffic emissions and traffic safety.Therefore,this paper focuses on the impact of heterogeneity of driving behavior on fuel consumption and traffic safety.Firstly,the I-80 section of NGSIM data set is chosen as the basic research data.Symmetric exponential moving average filtering algorithm(SMEA)is applied to filter the noise of raw data.According to the different data requirements,car-following and lane-changing driving trajectory are extracted from the processed data set.Secondly,in order to study the heterogeneity of car-following behavior,FVD,Gipps,IDM car-following models are selected to simulate the car-following behavior.Genetic algorithm is applied to calibrate the parameters of car-following models at an individual vehicle’s level.Principal component analysis and K-means clustering method are further used to cluster driving styles.According to the clustering results,the driving style of car-following behavior could be divided into three types: aggressive,conservative and moderate.Each of driving style can be characterized by different car-following model parameters.Then,starting from the analysis of the mechanism of lane-changing behavior,the main factors which could reflect heterogeneity of lane-changing behavior are selected.According to the clustering algorithm,the driving style of lane-changing behavior could be categorized into three types: aggressive,conservative and moderate.Based on the analysis of lane-changing process of vehicles,lane-changing rules considering different lane-changing styles are proposed.An improved Cellular automata model is proposed in this paper,which incorporated the heterogeneity of driving styles into the Na Sch model and the STCA model.Finally,different simulation experiments are carried out to study the heterogeneity of car-following behavior and lane-changing behavior through the MATLAB simulation platform,and traffic flow simulation scenes containing different ratios of driving styles are set up.In the simulation of the heterogeneity of car-following behavior,the fuel consumption and safety of the simulated road section are analyzed based on the comparative experiments of three different car-following models,combined with the emission model and the safety index model.The simulation results illustrate that as the ratio of conservative driving styles decreases or the proportion of aggressive driving styles increases in the traffic flow,fuel consumption in the transportation system has an upward trend to increase.At the same time,the increase in the ratio of moderate drivers will lead to reduce the level of traffic safety.Simulation research on the heterogeneity of lane-changing behavior shows that the increase of the proportion of aggressive driving style will not only increase fuel consumption,but also reduce the safety level of the transportation system.