VISSIM交通仿真模型参数校正及主辅路出入口设置研究

发布时间：2018-03-29 04:17

本文选题：交通仿真　切入点：VISSIM　出处：《首都经济贸易大学》2015年硕士论文

【摘要】：城市快速路的修建主要是为了解决交通需求的高速增长,在其与相交道路连接的必要过渡处即出入口,却成为造成城市快速路拥堵的一大原因。本文采用微观交通仿真方法,解决这一关键问题。交通仿真是模拟现实交通状况分析交通状态的有效方法,目前我国使用的交通仿真软件大多由国外研发,其模型参数设置不适于我国的交通流特性。因此本文将从模型参数校正出发,结合群智能优化算法的改进研究开发一个高效准确的参数校正程序,应用于城市快速路出入口仿真中。首先,对人工蜂群算法进行改进研究。人工蜂群算法(ABC)是模拟蜜蜂采蜜行为的群智能优化算法,针对其容易陷入早熟收敛等不足,引入文化算法中的双层进化结构和多种群并行进化思想,提出基于双层进化的多种群并行人工蜂群算法(PMABC)。将采蜜蜂群划分为具有不同搜索策略的子种群并行进化,平衡全局开发能力与局部搜索能力,避免算法过早陷入局部最优。采用双层进化结构,采蜜蜂群作为种群空间寻找可行解,追随蜂群作为信仰空间,记忆采蜜蜂群搜索的优质蜜源并继续搜索,其搜索结果用于指导蜂群寻优,加速算法收敛,提高收敛精度。通过六个经典的适应度测试函数仿真实验表明,该算法能够有效避免算法陷入局部最优并具有较快的收敛速度和较高的收敛精度。其次,结合VISSIM使用手册介绍VISSIM软件特点,说明选择其作为本次校正及仿真工具的原因。针对主辅路出入口这一研究对象,确定需要校正的参数和参数校正的评价指标,选择北京三环光华桥附近的出入口进行数据采集,建立仿真模型。通过COM接口使用VBA语言开发一套自动校正程序,该程序所使用的优化算法为多种群并行人工蜂群算法。最后,分析了校正结果,从而证明了本文所提出的改进的群智能优化算法和自动校正程序是有效的,极大的提高了模型的准确性,且多种群并行人工蜂群算法的运用提高了校正的速度。该校正程序可应用于各个微观交通仿真中,具有普适性,对于我国交通仿真的研究与应用具有重要意义。
[Abstract]:The construction of urban expressway is mainly to solve the rapid growth of traffic demand. The entrance and exit, which are necessary to connect with the intersecting road, have become a major cause of urban expressway congestion. This paper adopts microscopic traffic simulation method. Traffic simulation is an effective method to simulate the real traffic situation and analyze the traffic state. At present, most of the traffic simulation software used in our country are developed by foreign countries. The model parameter setting is not suitable for the traffic flow characteristics in China, so this paper will develop an efficient and accurate parameter correction program based on the model parameter correction and the improvement of the swarm intelligence optimization algorithm. First of all, the artificial bee colony algorithm is improved. The artificial bee colony algorithm (ABC) is a swarm intelligence optimization algorithm to simulate honeybee honey gathering behavior, which is easy to fall into premature convergence and other shortcomings. By introducing the two-layer evolutionary structure in the cultural algorithm and the idea of multi-population parallel evolution, a multi-population parallel artificial swarm algorithm based on double-layer evolution is proposed. The honeybee colony is divided into sub-populations with different search strategies. The global development ability and the local search ability are balanced to avoid the algorithm falling into local optimum prematurely. The two-layer evolutionary structure is adopted to find the feasible solution in the population space of honeycomb, and to follow the colony as the belief space. The high quality honey source of honeybee colony search is memorized and continued. The search results are used to guide the colony optimization, accelerate the convergence of the algorithm and improve the convergence accuracy. The simulation results of six classical fitness test functions show that, The algorithm can effectively avoid falling into local optimum, and has faster convergence speed and higher convergence precision. Secondly, the characteristics of VISSIM software are introduced with the help of VISSIM manual. Explain the reason why it is chosen as the tool of correction and simulation. In view of the main and auxiliary road entrance and exit, determine the parameters that need correction and the evaluation index of parameter correction, and select the entrance and exit near Beijing Sanhuan Guanghua Bridge to collect data. The simulation model is established. A set of automatic correction program is developed by using VBA language through COM interface. The optimization algorithm used in this program is multi-colony parallel artificial colony algorithm. Finally, the correction results are analyzed. It is proved that the improved swarm intelligence optimization algorithm and the automatic correction program proposed in this paper are effective and greatly improve the accuracy of the model. The application of multiple swarm parallel artificial bee swarm algorithm improves the speed of correction. This program can be used in various microscopic traffic simulation, which is universal and has great significance for the research and application of traffic simulation in our country.
【学位授予单位】：首都经济贸易大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U491

【参考文献】