受控路网交通状态完备信息条件研究及应用

发布时间：2018-10-31 19:46

【摘要】：随着机动车保有量的不断增加,交通拥挤俨然已经成为影响城市功能正常发挥和可持续发展的一个全局性问题。智能交通被公认为是有效缓解交通拥堵的最佳途径,伴随着传感器网络、物联网、信息物理融合系统、云计算、大数据等先进技术的发展,智能交通信息服务模式有了新的变化,这使交通网络的完全控制成为可能。现在关于交通信息供给模式的研究还是基于粗略、不完整控制的路网条件,与实际应用环境有差异。在受控路网条件下,为达到完全控制效果,交通信息的供给模式应首先支持准确的交通状态描述。故而本文旨在建立基于交通状态描述的信息模型,获得描述交通状态的完备信息条件,以指导交通信息采集系统的建设。本文的研究内容主要包括:从受控路网的交通信息特征出发,分析了交通状态识别与路网拥挤管理的信息需求,并对比分析现有的各种交通信息采集技术,以此作为指导信息采集系统建设的依据;综合考虑路段的信息采集需求、路段的网络结构特征及其对路网交通状态描述的影响,构建了路段信息服务水平的评价指标体系,并运用多属性决策理论来量化信息服务水平;引入粗糙集理论,对其中的基本概念赋予交通内涵,论证了运用该理论进行信息完备条件提取的合理性。面向交通状态识别,选取时空属性、交通流特征属性以及状态属性作为系统的属性集合,建立了对象交通状态信息的知识表达系统,在分析对比多种算法效果后,选择遗传算法进行数据离散化处理、属性约简,并根据约简结果提取出基于交通状态识别的完备信息条件;通过对交通检测器布局影响因素的分析,建立了以系统成本最低、数据的可靠性最大以及系统信息服务水平最高为目标函数,以OD覆盖原则和信息完备原则为约束条件的多目标优化模型。应用宽容分层序列法,并对宽容系数进行了适用性调配,保证本文多目标优化模型的求解的可行性与有效性,最后采用遗传算法完成模型求解。在算例分析中,以Nguyen-Dupuis路网作为研究对象,通过VISSIM二次开发技术完成路网数据的获取;采用Rosetta软件实现了路网完备信息条件的提取,获得该路网的完备信息条件集合为(速度,流量)、(道路编号,行程时间,速度)、(行程时间,速度,占有率)、(道路编号,速度,占有率)、(道路编号,速度,排队长),并利用提取的规则验证了完备信息条件的准确性;最后利用MATLAB编程设计了检测器的优化布局,证明了本文模型的有效性与实用性。
[Abstract]:With the increasing number of motor vehicles, traffic congestion has become an overall problem affecting the normal development of urban functions and sustainable development. Intelligent transportation is recognized as the best way to alleviate traffic congestion effectively. With the development of sensor network, Internet of things, information physical fusion system, cloud computing, big data and other advanced technologies, The mode of intelligent transportation information service has been changed, which makes it possible to control the traffic network completely. The research on traffic information supply mode is based on rough, incomplete control of road network conditions, which is different from the actual application environment. Under the condition of controlled road network, in order to achieve the complete control effect, the supply mode of traffic information should support the accurate description of traffic state first. Therefore, the purpose of this paper is to establish an information model based on traffic state description, and to obtain complete information conditions to describe traffic state, so as to guide the construction of traffic information collection system. The main research contents of this paper are as follows: based on the traffic information characteristics of the controlled road network, the information needs of traffic state identification and congestion management are analyzed, and the existing traffic information collection technologies are compared and analyzed. This is the basis for guiding the construction of information collection system. Considering the demand of information collection, the network structure characteristics and its influence on road network traffic state description, the evaluation index system of road information service level is constructed, and the multi-attribute decision theory is used to quantify the information service level. The rough set theory is introduced to give traffic connotation to the basic concepts, and the reasonableness of using the theory to extract the complete information condition is proved. For traffic state recognition, we select space-time attribute, traffic flow characteristic attribute and state attribute as the attribute set of the system, and set up the knowledge representation system of the object traffic state information. After analyzing and comparing the effects of various algorithms, Genetic algorithm is selected for data discretization, attribute reduction, and the complete information condition based on traffic state recognition is extracted according to the reduction result. Based on the analysis of the factors affecting the layout of the traffic detector, the objective functions are the lowest system cost, the highest reliability of the data and the highest level of system information service. A multi-objective optimization model with OD covering principle and information completeness principle as constraints. The tolerance hierarchical sequence method is applied and the tolerance coefficient is adjusted to ensure the feasibility and validity of the multi-objective optimization model in this paper. Finally genetic algorithm is used to solve the model. In the example analysis, taking the Nguyen-Dupuis road network as the research object, through the VISSIM secondary development technology to complete the road network data acquisition; The complete information condition of the road network is extracted by using Rosetta software. The complete information conditions of the network are (speed, flow), (road number, travel time, speed), (travel time, speed, occupancy rate). (road number, speed, occupancy), (road number, speed, platoon captain), and use the extracted rules to verify the accuracy of the complete information conditions; Finally, the optimal layout of the detector is designed by using MATLAB, which proves the validity and practicability of the model.
【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】