面向城市道路交通排放测算的VISSIM仿真模型优化研究

发布时间：2018-06-11 23:49

本文选题：Wiedemann74跟驰模型 + 排放测算　；参考：《西南交通大学》2017年硕士论文

【摘要】：随着城市空气污染问题日益严重,雾霾天气持续不断,城市机动车排放所导致的污染问题进一步地突显。为了进一步量化交通排放,最新版本的PTVVISSIM微观交通仿真软件已开发了排放测算的功能,但其测算的污染物类型单一且精度较低。将微观交通仿真模型与交通排放模型相结合对城市交通排放进行评价,仍然是当前相关领域的研究热点。由于VISSIM微观交通仿真软件的核心模型,Wiedemann74跟驰模型在进行排放测算时的存在系统误差。因此本文引入了机动车排放的最佳解释变量VSP(Vehicle Specific Power),基于大量实测城市道路驾驶行为数据,展开对跟驰模型的优化研究。本文分析了基于VSP分布的微观排放模型、VISSIM微观仿真模型参数校正及优化和面向交通排放VISSIM仿真模型优化的国内外研究现状。重点研究了 VSP变量和Wiedemann74跟驰模型的内部跟驰逻辑和仿真机理。通过SIMI软件分析研究路段实际交通流视频,得到了城市道路驾驶行为数据。在进行数据质量控制与处理后,展开了城市道路驾驶行为特性的分析。根据Wiedemann74跟驰模型参数的默认值设计了 MATLAB仿真程序,并对比分析了实测数据与仿真数据平均速度区间上的VSP分布。根据各行驶状态所占时间比例与VSP分布的相对均方根误差,得出不能通过调整速度跟驰阈值对排放测算精度进行优化。进一步确定了距离跟驰阈值的模型跟驰域优化方案,其优化效果明显,尤其对高速区间的VSP分布的优化效果显著。经过对VSP分布的相对均方根误差的敏感性分析,确定了最大加速度与最大减速度的优化的最佳方案。最大加、减速度优化能很大程度地降低低速区间与高速区间的VSP分布的相对均方根误差。Wiedemann74跟驰模型整体优化后各行驶状态的划分更加接近于车辆的实际行驶状态,随着平均速度的增加,整体优化后的模型与原Wiedemann74跟驰模型仿真VSP的相对均方根误差总体都呈现出上升的趋势。整体优化后的跟驰模型输出的VSP分布相对均方根误差降低50%以上。整体优化后的Wiedemann74跟驰模型在稳定性上与原跟驰跟驰模型保持一致。
[Abstract]:With the urban air pollution problem becoming more and more serious and the smog weather continuing, the pollution problem caused by urban motor vehicle emissions is becoming more and more prominent. In order to further quantify traffic emissions, the latest version of PTVVISSIM microcosmic traffic simulation software has developed the function of emission measurement, but the type of pollutants measured by PTVVISSIM is single and the accuracy is low. The combination of microscopic traffic simulation model and traffic emission model to evaluate urban traffic emissions is still a research hotspot in related fields. Because of the systematic error of Wiedemann74 model which is the core model of VISSIM microscopic traffic simulation software. Therefore, this paper introduces the best explanatory variable of vehicle emission, VSP vehicle specific Powertrain, based on a large number of measured urban road driving behavior data, the optimization of the car-following model is carried out. In this paper, the research status of VSP distributed micro emission model (VISSIM) parameters correction and optimization and traffic emission VISSIM simulation model optimization at home and abroad is analyzed. The internal car-following logic and simulation mechanism of VSP variable and Wiedemann74 car-following model are studied. The actual traffic flow video is analyzed by SIMI software, and the data of urban road driving behavior are obtained. After data quality control and processing, the characteristics of urban road driving behavior are analyzed. According to the default values of Wiedemann74 model parameters, a MATLAB simulation program is designed, and the VSP distribution on the average velocity range of measured and simulated data is compared and analyzed. According to the RMS error between the time ratio of each driving state and the VSP distribution, it is concluded that the accuracy of emission measurement can not be optimized by adjusting the speed and the car-threshold. The optimization scheme of the model followed by the range and car-following threshold is further determined, and the optimization effect is obvious, especially for the VSP distribution in the high-speed region. Based on the sensitivity analysis of the relative root mean square error of VSP distribution, the optimum scheme of maximum acceleration and maximum deceleration is determined. The maximum addition and deceleration optimization can greatly reduce the relative root mean square error (RMSE) of the VSP distribution in the low speed range and the high speed interval. The partition of each driving state is closer to the actual driving state of the vehicle after the overall optimization of the model. With the increase of average velocity, the relative root mean square error (RMS) of the whole optimized model and the original Wiedemann74 model is increasing. After the overall optimization, the VSP distribution of the car-following model is reduced by more than 50% relative to the root mean square error (RMS). The overall optimized Wiedemann74 car-following model is consistent with the original car-following model in terms of stability.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491

【参考文献】