Study On Emission Analysis And Simulation Optimization Considering Driving Behavior

With the rapid increase of car ownership,the problem of vehicle exhaust pollution has received extensive attention in the past decade.As an important means of connecting the urban traffic flow evaluation with vehicle exhaust emission calculating,the microscopic traffic simulation model has been studied by many scholars.However,there is currently little research focusing on the influence of driver behavior characteristics on vehicle exhaust emissions and microscopic traffic simulation modeling.Moreover,although the latest version of the microscopic traffic simulation software VISSIM is already equipped with the emission measurement function,the accuracy of emission measurement can not meet the needs of the actual traffic flow evaluation.Therefore,this thesis analyzes the exhaust emission measurement of motor vehicles based on the characteristics of driving behavior.With the goal of improving the accuracy of emission estimation.This paper also optimizes car following model Wiedemann 74 model,which is the core model of the VISSIM software,and the effect of optimization was verified.Firstly,follow-up surveys and questionnaire surveys were conducted by 28 drivers with different backgrounds and driving experiences in Chengdu.In the process of survey,the trajectory data of the investigated vehicles were recorded through GPS devices which are installed on the survey vehicles previously.A data processing model was constructed and data preprocessing was performed based on the principle of Kalman filtering.Secondly,drivers are grouped into four categories based on observations of the driver’s behavior characteristics such as acceleration in the low-speed phase and deceleration in the high-speed phase,and self-assessment result in the driver’s driving behavior questionnaire,including radical or conservative levels,driving experience,etc.At the same time,the characteristics of the car-following parameters of the four types of drivers as well as the saturation flow rate characteristics of different types of drivers are further studied.The results show that cruising speeds vary from different types of drivers.For example,the saturation flow rates of novice,female,and conservative drivers are lower than other types of drivers.On this basis,the emission analysis of 28 drivers involved in the survey was conducted with the performance index of vehicle specific power VSP.For all types of drivers,the VSP distribution,the main pollutants emitted by the vehicle,and the fuel consumption were studied.The results show that the VSP distribution of Type 1 and Type 3 drivers are similar,and that of Type 2 and Type 4 drivers are similar.The results show that the CO,NOx,and HC emissions of Type 1 drivers are significantly higher than other types of drivers.Type 3 drivers use the least amount of fuel on the journey,while Type 2 drivers use the most.Finally,according to the construction principle and logic of the Wiedemann74 car-following model,the numerical simulation method of the model is designed using Matlab.By combining with the built-in default value of the model in VISSIM,the simulation output based on the default parameter value turns out to have a large difference between measured data.Besides,this thesis uses the measured data to optimize the Wiedemann74 model.Through the study of the car-following threshold,maximum acceleration and maximum deceleration of all types of drivers in the model,we propose the optimization scheme for different types of drivers with the aim of increasing the accuracy of model’s emission measurement.The effect of the optimization scheme is evaluated,turns out that the model simulation output of each type of driver is very close to the measured data,the root mean square error of VSP is significantly reduced,and various types of driving are verified.The simulated VSP distribution and the measured VSP of four types of drivers all follow the same distribution.Therefore,the conclusion drawn from this thesis is that the overall optimization scheme has a significant impact on improving the measure accuracy of the simulating emission by Wiedemann74 car-following model,which is obviously effective.