【作者】 王鹏；

【导师】 刘卫宁；

【作者基本信息】 重庆大学 ， 交通信息工程及控制， 2010， 硕士

【摘要】 车辆超速行驶是导致交通事故的主要原因之一,如何对车辆超速进行科学的管理是管理部门面临的重要问题。论文在讨论车辆超速监测与分析国内外研究现状的基础上,指出目前对于车辆超速数据的采集主要依靠人力和昂贵的硬件设备,对于车辆超速的分析主要依靠经验,缺乏充分的数据和先进技术的支持,无法对超速行为进行多样化、多层次、多角度的分析,管理决策往往缺乏针对性和有效性。针对上述问题,论文提出利用现有车辆上的GPS(Global Positioning System,全球定位系统)定位装置,依靠交通管理部门的GPS监控平台获取的车辆行驶数据进行车辆超速规律分析,支持对车辆超速问题的针对性管理。论文将GPS数据作为分析车辆超速的依据,以超速数据采集处理、整体解决方案的构建、超速数据仓库模型研究、OLAP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GPS超速报警数据,利用最近点估计地图匹配技术,对超速数据进行道路匹配的预处理。结合目前道路运输安全管理部门对营运车辆超速规律分析的需求,本着合理性、稳定性、经济性、可操作性和可维护性等方面的设计思想,提出了营运车辆超速规律多维分析的总体解决方案。针对总体方案中需要解决的关键问题进行了深入研究。首先,根据目前道路运输管理部门的数据环境及数据特点,研究了基于DTS的数据抽取器。其次,对超速规律分析的数据仓库模型和OLAP数据库进行了研究,重点分析了超速数据的粒度,并对超速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–雪花模型并完成了物理实现。基于上述数据仓库模型,选取道路、企业、时间作为超速分析多维数据集的维度,创建了超速分析多维数据集。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,通过对现有的多维数据集更新方式进行改进,提出了基于最优时间判别模型的多维数据集更新方法。应用上述研究成果,利用重庆市GPS监控平台采集的车辆报警数据,开发了重庆市营运车辆超速规律多维分析系统,为提高重庆市运输安全管理决策的科学性起到了重要的支撑作用。更多还原

【Abstract】 Speeding is one of the main causes of traffic accidents. How to make scientific management of vehicle speed is an important problem facing the management department. Based on analyzing the current research status of vehicle speed, it is pointed out that rely mainly on human and expensive hardware equipments for vehicle speeding data acquisition. The analysis for vehicle speed relies mainly on the experience, lack of sufficient data and advanced technical support. So managers cannot to conduct diversification, multi-level and multi-angle analysis, lead to the decision lacks pertinence and effectiveness.According to the above-mentioned problems, the paper provide basis for vehicle speed analysis based on the GPS device and GPS data, in order to realize the targeted for vehicle speed auxiliary management.GPS data as papers will be based on analysis of vehicle speed. Paper’s research clues are speeding data acquisition and processing, the overall solutions, speeding data warehouse model research, OLAP multi-dimension data design, data warehouse and multi-dimension data updates. For speeding alarm data, the paper realizing speeding data pretreatment use nearest point estimated map matching. Combining the demand of road transportation security administration departments for speeding analysis, a general solution to multidimensional analysis of vehicle speeding is presented according to the design thought of economics, rationality, stability, maneuverability and maintainability.Aiming at the key problems which need to be solved in the general solution, a universal data extractor is designed in the article in terms of the current data environment and data traits of the transportation management sectors. The paper researched the model of data warehouse and OLAP database, including the study on the speeding data granularity, speeding data based on granularity of generalized, choose the minimal granularity as data grand. Then, the star-snowflake model of the data warehouse is constructed and physically realized.Based on the data warehouse model, a cube was created for speeding analysis, and the road, enterprise and time were choosing as the dimension of cube. Finally, the practical application of the environment, the paper studied the method and procedure of data warehouse update, then improved the existing update method of multi-dimension. This paper puts forward the multi-dimension data cube updating method based on optimal time model.Finally, OLAP system for Chongqing City’s roadway transportation safety management is developed and realized with the application of the achievements in the above research, using the vehicle alarm data collected by the GPS monitoring system in Chongqing. This system is of significance in supporting the improvement of the scientific in actual transportation management and decision-making.更多还原