基于手机轨迹数据的用户出行及停驻点识别系统研究

发布时间：2019-01-21 06:50

【摘要】：移动手机轨迹数据由于其样本量大、实时性强并且能够在一定程度上描述手机用户的出行行为而正得到研究人员的重视，并被广泛地应用于城市计算。本文主要围绕移动手机轨迹数据开展了以下几方面的工作： 1.开发了“基于语义技术的手机定位识别系统”。该系统的主要目的是基于手机轨迹数据分析用户出行的起讫点和出行目的。采用了语义技术来描述多源的交通数据所包含的语义信息，继而建立了交通数据的语义表示数据库。在该系统中，可以通过设置规则来进行逻辑推理，从而实现了使用手机轨迹数据来分析用户出行目的功能。 2.由于对手机用户隐私保护的需要，研究人员很难获取到移动手机轨迹数据。同时，，即使能够获取到移动手机轨迹数据，这些数据也由于缺少语义标注而不利于科学研究的开展。因此，有必要根据手机轨迹数据定位产生的通信原理对手机轨迹数据进行仿真以更好地支持对手机轨迹数据的研究及应用工作。为此，本文搭建了一个“基于用户行为的移动数据仿真系统”。在系统中首先设定了一个接近真实的移动通信仿真系统；接着，在用户真实出行活动的描述基础上，对用户出行活动中的轨迹点进行了提取和语义标注；然后，在仿真的移动通信系统中复现用户出行的过程，并记录复现过程中用户的通信事件信息；最后，生成带有语义标注的手机轨迹数据。该系统在活动情景仿真、人的行为仿真、移动通信系统仿真方面都具有优势。 3.手机用户停驻点识别是利用移动手机轨迹数据分析用户出行活动的关键步骤。针对现有停驻点识别中存在的问题，将手机用户轨迹与背景地理语义信息产生的功能区相结合，通过考虑用户在某一功能区域内的整体活动来判断用户是否在此停驻。使用本方法判断的停驻点，不仅仅是限于一个小的范围（单个小区的覆盖范围），而是由多个相邻且功能相似小区组成的功能区中的小区集合。通过对带有标签的仿真的手机轨迹数据进行实验，证明了新的停驻点识别方法有效地避免了由于正常切换、错误切换带来的误判。
[Abstract]:Mobile phone trajectory data is being paid attention by researchers because of its large sample size, real-time ability and the ability to describe the mobile phone user's travel behavior to a certain extent, and has been widely used in urban computing. This paper mainly focuses on mobile phone trajectory data in the following areas: 1. A mobile phone location recognition system based on semantic technology is developed. The main purpose of this system is to analyze the starting point and destination of user travel based on mobile phone trajectory data. Semantic technology is used to describe the semantic information contained in multi-source traffic data, and then the semantic representation database of traffic data is established. In this system, we can use the mobile phone trajectory data to analyze the user travel purpose function by setting the rules to carry out logical reasoning. 2. Due to the need of privacy protection for mobile phone users, it is difficult for researchers to obtain mobile phone trajectory data. At the same time, even if mobile phone trajectory data can be obtained, these data are not conducive to scientific research due to the lack of semantic tagging. Therefore, it is necessary to simulate the mobile phone trajectory data according to the communication principle of mobile phone locus data location to better support the research and application of mobile phone trajectory data. Therefore, a mobile data simulation system based on user behavior is built in this paper. In the system, a close to real mobile communication simulation system is first set up, and then, on the basis of the description of the user's real travel activity, the trace points in the user's travel activity are extracted and the semantic tagging is carried out. Then, the user travel process is reproduced in the simulated mobile communication system, and the communication event information of the user is recorded. Finally, the mobile phone trajectory data with semantic annotation is generated. The system has advantages in activity scenario simulation, human behavior simulation and mobile communication system simulation. 3. The identification of stop point of mobile phone user is a key step to analyze user travel activity using mobile phone trajectory data. In view of the problems existing in the existing stop-point identification, this paper combines the mobile phone user track with the functional area generated by the background geographic semantic information, and determines whether the user is staying here by considering the overall activities of the user in a certain functional area. The stopping point determined by this method is not only limited to a small range (the coverage of a single cell), but also a set of cells in a functional area consisting of several adjacent and functionally similar cells. Through the experiment of the simulated mobile phone trajectory data with labels, it is proved that the new method can effectively avoid the misjudgment caused by the normal switching and wrong switching.
【学位授予单位】：北京工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN929.53

【共引文献】