Optimization On VISSIM Simulation Model For The Evaluation Of Traffic Emissions On Urban Roads

With the increasingly serious problem of urban air pollution,fog and haze continued and the pollution problem caused by urban vehicle emissions has been further highlights.In order to further quantify the traffic emissions,the latest version of the PTV VIS SIM microscopic traffic simulation software has been developed a function to evaluate the emissions,but its measured pollutant type is simple and low accuracy.It is still a hot spot in the research field to evaluate the urban traffic emissions by combining the microscopic traffic simulation model and traffic emission model.The car-following model,Wiedemann74 is the core model for VISSIM,but according to some researches,this model can have systematic errors when it has been applied to evaluate the emissions.Therefore,this paper uses a large amount of real vehicle behavior data to conduct a research on the improvement of this car-following model with the help of the variable VSP(Vehicle Specific Power),which is the best explanatory variable of vehicle emissions.This paper has reviewed a wide variety of researches on microscopic emission models based on VSP distribution,the parameter calibration and optimization of VISSIM microscopic simulation model as well as the optimization of the VISSIM connected with traffic emissions.What's more,this paper focuses on the variable of VSP and the following logic and simulation mechanism of Wiedemann74 car-following model.Based on the analysis of the actual traffic flow video by SIMI software,the driving behavior data of urban road are obtained.The research on the characteristics of urban road driving behavior were further conducted after the data quality control and processing.The MATLAB simulation program is designed according to the default value of the Wiedemann74 car-following model parameters,and the VSP distribution on the average velocity interval between the measured data and the simulation data is compared and analyzed.According to the proportion of the running time and the relative root mean square error of VSP distribution,it is found that the accuracy of the emission evaluation cannot be optimized by adjusting the threshold of speed in the car-following model.This result has further driven the research to the direction to adjust the threshold of distance in the car-following model.This new results were turned out to be more effective in terms of the optimization of the model,especially for the optimization of the VSP distribution in the high-speed section.Through the sensitivity analysis of the relative root mean square error of the VSP distribution,the optimal scheme of the maximum acceleration and the maximum deceleration was finally determined.This scheme has reduced the relative-root-mean-square error of the VSP distribution in both low-speed and high speed section to a large extent.The new characterization of driving behaviors based on Wiedemann74 car-following model is more realistic.Although the relative root mean square error of VSP distribution simulated by optimized model and the original Wiedemann74 has shown a common increasing trend with the augmentation of the average speed.Compared with the original car-following model,the relative-root-mean-square error of VSP distribution simulated by the optimized model has been reduced more than 50%,while these two models has presented a common level of stability.