《2020（交通运输）基于高速公路收费数据的交通流量多维分析技术研究》由会员分享，可在线阅读，更多相关《2020（交通运输）基于高速公路收费数据的交通流量多维分析技术研究（67页珍藏版）》请在金锄头文库上搜索。
1、（交通运输）基于高速公路收费数据的交通流量多维分析技术研究基于高速公路收费数据的交通流量多维分析技术研究重庆大学硕士学位论文学生姓名：万金朋导师姓名：孙棣华 教 授专 业：交通信息工程及控制学科门类：工 学重庆大学自动化学院二O一二年四月Study on Multidimensional Analysis of Traffic Flow based on Highway Toll Data A Thesis Submitted to Chongqing Universityin Partial Fulfillment of the Requirement for theDegree of Ma
2、ster of EngineeringByWan Jinpeng Supervised by Prof. Sun DihuaMajor: Transportation Information Engineering and ControlCollege of Automation of Chongqing University, Chongqing, ChinaApril, 2012摘 要交通流量是反映交通运行状态的主要参数之一，如何科学的分析高速公路的交通流量分布规律是高速公路管理部门面临的重要问题之一。论文在讨论交通流量的获取与交通流量分析的国内外研究现状的基础上，指出目前对于交通流量的采
3、集主要依靠人力或昂贵的硬件设备，高速公路的交通流量很不全面，对于高速公路流量的分析主要依靠经验，缺乏充分的数据和先进技术的支持，无法对交通流量进行多样化、多层次、多角度的分析，管理决策往往缺乏针对性和有效性等缺点。针对上述问题，论文利用现有的高速公路联网收费系统中的收费站出入口数据，提出了一种高速公路断面的交通流量的估算算法，进而采用趋势图、透视表等数据可视化方法对高速公路不同时间、断面以及车型的交通流量数据进行分析，获取其时空分布规律，加深对交通流量数据的理解，有助于进行高速公路交通流量的时空数据挖掘的研究，为有关部门提供科学管理和决策支持工具，支持对高速公路的针对性管理。论文将联网收费系统
4、的收费数据作为分析高速公路断面的交通流量的依据，以整体框架的构建、交通流量的估算处理、相关信息的抽取、交通流量仓库模型研究、OLAP(On Line Analytical Processing)多维数据集的设计、数据仓库和多维数据集的更新为线索展开研究，对于交通流量分析结果，利用OWC(Office Web Component)组件进行多维展示。针对总体框架中需要解决的关键问题进行了深入研究。首先，根据目前高速公路管理部门的数据环境及数据特点，研究了基于DTS（Data Transformation Services）的交通流量数据抽取器，进而提出了一种基于高速公路联网收费数据的多断面交通流量
5、估算方法。其次，对交通流量规律分析的数据仓库模型和OLAP数据库进行了研究，重点分析了交通流量数据的粒度，并对交通流量数据进行了基于粒度的概化，在此基础上，构造了数据仓库的星型模型并完成了物理实现。基于上述数据仓库模型，选取断面、车型、时间作为交通流量分析多维数据集的维度，创建了交通流量分析多维数据集。最后，针对实际应用环境的需要，研究了数据仓库更新的方法和流程。应用上述研究成果，利用重庆市联网收费平台采集的高速公路收费数据，开发了重庆市高速公路断面交通流量规律多维分析系统，具有良好的应用前景。关键词：交通流量分析，收费数据，OLAP，交通流量数据仓库，数据更新ABSTRACTTraffic
6、flow is one of the main parameters to reflect the state of traffic running, How the scientific analysis of highway traffic distribution rule is one of the important issues facing the highway management department. The Paper to discuss traffic flow on the basis of the research status and traffic flow
7、 analysis, which points out human or expensive hardware for the acquisition of the traffic flow. Highway traffic flow is far from comprehensive, which mainly replys on experience in the analysis of freeway traffic. Lack of adequate data and advanced technology support, which can not be on the traffi
8、c flow, diversified, multi-layered, multi-angle analysis. Management decisions often lack relevance and effectiveness of other shortcomings.In response to these issues, The Paper takes use of the toll station entrances and exits in the data in the existing Highway Toll System and makes a point of a
9、highway sections traffic flow estimation algorithm highway sections traffic flow estimation algorithm. Thus it adopts the trends, pivot tables and other data visualization methods,which rsponds to the analysis of the highway at different times, the cross section as well as models of traffic flow dat
10、a and access to its spatial and temporal distribution, deeply understands the traffic flow data. It contributes to the highway traffic flow spatio-temporal data mining research,and provides scientific management and decision support tools for the relevant departments. Support targeted on the problem
11、 of highway traffic flow management. The Paper Toll System charges data as a basis for analysis of traffic flow on the highway section and uses the construction of the overall framework,traffic flow estimation processing, the extraction of relevant information, the design of the OLAP cube, data ware
12、house and cube updates provide some clues. For traffic flow analysis, it takes advantage of OWC components of the multi-dimensional to show.Key issues need to be addressed in the overall framework of the in-depth study. First of all, according to the data environment and data characteristics of the
13、highway management department, it researchs the traffic flow data extraction based on DTS,and points out a traffic flow estimation method based on the data section of the Expressway Network charges.Secondly,it poses model law analysis of traffic flow data warehouse and OLAP databases,and mainly anal
14、yzes the traffic flow data granularity,and the traffic flow data based on the granularity of generalizability.On this basis, it constructes star model of the data warehouse and completes the physical implementation.Based on the above data warehouse model, it selects the section、models、times to make up of the dimension of the cube as
