VISSIM仿真参数校正的宏观方法研究
发布时间：2018-07-29 20:36
【摘要】：目前交通问题日益严峻,研究者们开始致力于提高现有道路、交叉口通行能力的研究来试图缓解城市交通拥挤。在研究这些问题时不可避免地应用到了交通仿真。交通仿真模型中有大量的参数,对仿真模型的准确性有较大影响。常用的仿真软件多是由国外开发,通常应用软件模型中的默认参数,因此确定适合我国道路交通的具体模型参数亟待解决,本论文针对微观仿真模型参数校正问题展开研究。文献综述介绍了国内外有关微观交通仿真模型参数校正的研究情况,然后介绍了微观交通仿真模型参数的校正流程、选取的评价指标以及采用的寻优算法,最后根据实测数据给出了Wiedemann跟车模型中的参数校正结果及车辆驾驶行为阈值的表示方法。由于流量速度图形能够表示自由流、拥堵和排队消散这三种交通状态,它所具有的信息量比较完整,因此采用流量速度图作为微观交通仿真模型参数的校正的评价指标。采用模式识别领域中的差异性度量方法,计算仿真输出的数据形成的流量速度图形和实际现场测得的速度流量图形之间的差异值大小。分别对高速公路模型(Wiedemann99模型)和城市道路模型(Wiedemann74模型)的仿真参数校正进行了研究。最后,结合遗传算法建立了VISSIM、MATLAB和ExcelLink集成平台,实现VISSIM参数的自动化校正过程。针对Wiedemann的两种不同模型,分别对高速公路、城市道路进行了数据采集,并应用开发的仿真参数自动校正系统平台对两个模型的参数进行校正,说明在仿真过程中模型参数校正的必要性。依据实际测得的长沙市南二环交通数据,对典型的Wiedemann车辆跟车行驶阈值进行拟合,并得到实际交通运行中的车辆行驶阈值图。
[Abstract]:At present, the traffic problem is becoming more and more serious, and researchers are trying to improve the capacity of existing roads and intersections to alleviate urban traffic congestion. In studying these problems, traffic simulation is inevitably applied. There are a lot of parameters in the traffic simulation model, which have great influence on the accuracy of the simulation model. Most of the commonly used simulation software are developed by foreign countries, and usually the default parameters in the software model are applied. Therefore, it is urgent to determine the specific model parameters suitable for road traffic in China. In this paper, the problem of parameter correction of microscopic simulation model is studied. This paper introduces the domestic and international research on the parameter correction of microscopic traffic simulation model, and then introduces the calibration flow, the selected evaluation index and the optimization algorithm used in the microscopic traffic simulation model. Finally, according to the measured data, the correction results of the parameters in the Wiedemann model and the representation method of the threshold value of the vehicle driving behavior are given. Because the flow velocity graph can represent the three traffic states of free flow, congestion and queue dissipation, it has relatively complete information, so the flow velocity graph is used as the evaluation index for the correction of the parameters of the microscopic traffic simulation model. Using the difference measurement method in the field of pattern recognition, the magnitude of the difference between the flow velocity graph formed by the simulated output data and the velocity flow figure measured in the actual field is calculated. The emulation parameter correction of highway model (Wiedemann99 model) and urban road model (Wiedemann74 model) are studied respectively. Finally, the integrated platform of Visual Simma MATLAB and ExcelLink is established based on genetic algorithm to realize the automatic correction of VISSIM parameters. According to two different models of Wiedemann, the data of expressway and urban road are collected, and the parameters of the two models are corrected by using the developed simulation parameter automatic correction system platform. The necessity of model parameter correction in the process of simulation is explained. According to the traffic data of Nanerhuan in Changsha City, the typical Wiedemann vehicle driving threshold is fitted, and the vehicle driving threshold map in actual traffic operation is obtained.
【学位授予单位】：长沙理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491
，
本文编号：2153941
[Abstract]:At present, the traffic problem is becoming more and more serious, and researchers are trying to improve the capacity of existing roads and intersections to alleviate urban traffic congestion. In studying these problems, traffic simulation is inevitably applied. There are a lot of parameters in the traffic simulation model, which have great influence on the accuracy of the simulation model. Most of the commonly used simulation software are developed by foreign countries, and usually the default parameters in the software model are applied. Therefore, it is urgent to determine the specific model parameters suitable for road traffic in China. In this paper, the problem of parameter correction of microscopic simulation model is studied. This paper introduces the domestic and international research on the parameter correction of microscopic traffic simulation model, and then introduces the calibration flow, the selected evaluation index and the optimization algorithm used in the microscopic traffic simulation model. Finally, according to the measured data, the correction results of the parameters in the Wiedemann model and the representation method of the threshold value of the vehicle driving behavior are given. Because the flow velocity graph can represent the three traffic states of free flow, congestion and queue dissipation, it has relatively complete information, so the flow velocity graph is used as the evaluation index for the correction of the parameters of the microscopic traffic simulation model. Using the difference measurement method in the field of pattern recognition, the magnitude of the difference between the flow velocity graph formed by the simulated output data and the velocity flow figure measured in the actual field is calculated. The emulation parameter correction of highway model (Wiedemann99 model) and urban road model (Wiedemann74 model) are studied respectively. Finally, the integrated platform of Visual Simma MATLAB and ExcelLink is established based on genetic algorithm to realize the automatic correction of VISSIM parameters. According to two different models of Wiedemann, the data of expressway and urban road are collected, and the parameters of the two models are corrected by using the developed simulation parameter automatic correction system platform. The necessity of model parameter correction in the process of simulation is explained. According to the traffic data of Nanerhuan in Changsha City, the typical Wiedemann vehicle driving threshold is fitted, and the vehicle driving threshold map in actual traffic operation is obtained.
【学位授予单位】：长沙理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491
，
本文编号：2153941
本文链接：https://www.wllwen.com/kejilunwen/jiaotonggongchenglunwen/2153941.html
