基于LBS的车辆出行时空特征研究

发布时间：2018-06-16 10:14

本文选题：时空数据存储 + 行程识别　；参考：《武汉理工大学》2015年硕士论文

【摘要】：研究物流车辆出行的热点区域、时空特征和演变规律,有利于延伸物流车辆监管数据的获取生成、集成管理、共享服务和应用的信息服务链,带动基于车联网的移动车辆监管信息增值服务的高速增长,提升城市的智慧物流监管水平和基于车联网的物流运输管理服务能力,提高工业运输效率和运输商品安全,进一步助力城市的和谐和可持续发展。本文基于湖北省科技支撑计划项目《基于车联网的物流车辆协同监管关键技术及应用示范》(2014BAA146)的课题背景,研究了车辆GPS数据的采集、预处理方法,同时为了提高时空数据检索效率,设计了一种基于MongoDB的时空数据存储方案,提出了一种基于车辆速度和ACC状态的行程识别方法,并基于行程数据对车辆出行的时空特征进行了研究。本文的主要工作包括:(1)设计了基于MongoDB的时空数据存储方法。在研究过程中,通过对采集的GPS元数据进行数值编码格式化处理进行存储,缓解了磁盘空间的存储压力。同时,采用基于sharding的MongoDB集群架构进行时空数据存储,并建立了复合索引和地理空间索引,提升了时空数据检索效率。(2)提出了基于车辆速度和ACC状态的行程识别方法。根据Stop-Move模型的基本原理,研究了基于车速和ACC状态的轨迹点分类方法,并通过不断地对GPS数据流中的轨迹点进行分类和合并,实现了对车辆行程活动的识别,为研究车辆出行热点和时空特征提供了数据基础。(3)设计了基于LBS的出行时空特征提取方法,并分析了武汉市物流车辆的出行时空特征。首先根据车辆出行的行程特征设计了基于LBS的车辆出行偏好、热点区域等特征的提取方法,以及车辆热点区域吸引力、作息活动相似度的计算方法,实现了对车辆出行热点区域吸引力计算和作息规律程度的评价,并分析了武汉市物流车辆的出行时空特征,为基于车联网的物流运营管理提供了决策支持。本文的主要贡献在于结合特定的课题背景,对车辆出行时空特征分析的关键技术进行了研究和应用,能够为智慧物流领域的车辆轨迹数据挖掘研究和应用提供一定的理论参考和技术参考。
[Abstract]:The study of hot spots, space-time characteristics and evolution rules of logistics vehicle travel is conducive to extending the information service chain of logistics vehicle supervision data acquisition, integration management, sharing services and applications. To promote the rapid growth of mobile vehicle regulatory information value-added services based on vehicle networking, to enhance the level of intelligent logistics supervision in cities and the capacity of logistics and transportation management services based on vehicle networking, to improve the efficiency of industrial transport and transport commodity safety, Further contribute to the harmonious and sustainable development of the city. Based on the project of Hubei province science and technology support project < key technology and application demonstration of logistics and vehicle collaborative supervision based on vehicle networking "2014BAA146), this paper studies the acquisition and preprocessing method of vehicle GPS data. In order to improve the efficiency of spatio-temporal data retrieval, a spatio-temporal data storage scheme based on MongoDB is designed, and a travel recognition method based on vehicle speed and ACC state is proposed. The temporal and spatial characteristics of vehicle travel are studied based on travel data. The main work of this paper is to design a spatio-temporal data storage method based on MongoDB. In the research process, the storage pressure of the disk space is alleviated by the numerical coding and formatting of the collected GPS metadata. At the same time, MongoDB cluster architecture based on sharding is used to store spatio-temporal data, and the composite index and geospatial index are established to improve the efficiency of spatio-temporal data retrieval. According to the basic principle of Stop-Move model, the track point classification method based on speed and ACC state is studied, and the recognition of vehicle travel activity is realized by continuously classifying and merging the trajectory points in GPS data stream. This paper provides a data basis for the study of vehicle travel hot spots and space-time features. (3) A method for extracting travel space-time features based on LBS is designed, and the spatio-temporal characteristics of logistics vehicles in Wuhan are analyzed. Firstly, according to the travel characteristics of vehicles, the paper designs a LBS based method to extract the features of vehicle travel preferences, hot spots and so on, as well as the calculation method of the similarity between the attraction of vehicle hot spots and the activity of work and rest. This paper realizes the calculation of the attraction of the hot spot area of vehicle travel and the evaluation of the degree of the regularity of work and rest, and analyzes the travel space-time characteristics of the logistics vehicle in Wuhan, which provides decision support for the logistics operation management based on the vehicle network. The main contribution of this paper lies in the research and application of the key technology of vehicle travel space-time characteristic analysis combined with the specific subject background. It can provide some theoretical and technical reference for the research and application of vehicle trajectory data mining in the field of intelligent logistics.
【学位授予单位】：武汉理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U495

【参考文献】