基于GPS浮动车数据的城市居民活动规律研究

发布时间：2018-05-07 10:48

本文选题：GPS浮动车 + 地图匹配　；参考：《昆明理工大学》2017年硕士论文

【摘要】：随着社会不断发展,城市居民生活水平不断提高,城市居民活动更加频繁,城市居民作为城市的重要组成部分,研究城市居民活动规律对城市规划具有重要意义。传统采用问卷调查方式获取城市居民信息的方法具有一定局限性,近年来,人们通过在浮动车上安装GPS设备来获取城市居民活动信息,GPS定位数据记录了居民出行时间、出行地理位置等信息,为研究城市居民活动规律提供了丰富的数据资源,论文利用GPS浮动车数据对昆明市城市居民活动规律进行了研究。首先,数据预处理和地图匹配。论文介绍了 GPS定位原理以及GPS定位数据结构,并对数据进行了预处理和地图匹配,由于数据量过大,匹配过程较慢,为此论文采用缓冲区分析与几何匹配结合的方法将GPS定位数据匹配到相应的道路上,结果表明该方法匹配效率较高,效果较好;其次,乘客上下车点及出行轨迹点提取。浮动车载客状态是研究城市居民活动规律的重要参数,论文利用出租车载客状态,通过编写C#程序提取出了乘客上下车点和出行轨迹点,并利用统计分析的方法对城市居民出行行程长度和行程时间进行统计分析。最后,城市居民活动热点区域研究。论文在总结常用空间聚类算法的基础上,采用模糊C-均值(FCM)聚类算法方法对城市居民活动规律进行研究,该算法具有效率高,聚类速度快的优点,但模糊C-均值聚类算法是一种局部最优算法,并且聚类数目具有不确定性。针对上述缺点,论文提出了一种改进的(FCM)聚类算法,该算法采用自适应距离范数,给每一个簇添加自己的范数诱导矩阵,在聚类过程中通过引入分区系数(PC)、分类熵(CE)和XB指数三个指标来确定最佳聚类数目,实验结果表明了改进的模糊C-均值聚类算法可有效挖掘出城市居民活动热点区域。
[Abstract]:With the development of society, the living standard of urban residents is improving, and the activities of urban residents are more frequent. As an important part of the city, it is of great significance to study the law of urban residents' activities. The traditional method of obtaining information of urban residents by questionnaire survey has some limitations. In recent years, people have recorded the travel time of urban residents by installing GPS equipment in floating vehicles to obtain the information of urban residents' activities and GPS positioning data. Travel location and other information provide rich data resources for the study of urban residents' activities. This paper makes use of the GPS floating vehicle data to study the urban residents' activities in Kunming. First, data preprocessing and map matching. This paper introduces the principle of GPS positioning and the structure of GPS positioning data, and carries out the data preprocessing and map matching. Because of the large amount of data, the matching process is slow. In this paper, the method of combining buffer analysis with geometric matching is used to match the GPS location data to the corresponding roads. The results show that the method is more efficient and effective. Secondly, the point of passengers getting on and off and the point of travel trajectory are extracted. The state of floating vehicle is an important parameter to study the law of urban residents' activities. In this paper, by writing a C # program, we extract the points of passengers' embarkation and disembarkation and the point of travel trajectory by using the state of taxi carrying passengers. The travel length and travel time of urban residents are statistically analyzed by means of statistical analysis. Finally, the study on the hot spots of urban residents' activities. On the basis of summarizing the common spatial clustering algorithms, the fuzzy C- mean FCM-based clustering algorithm is used to study the urban residents' activity law. The algorithm has the advantages of high efficiency and fast clustering speed. But fuzzy C-means clustering algorithm is a kind of local optimal algorithm, and the number of clustering is uncertain. In order to overcome these shortcomings, an improved clustering algorithm is proposed in this paper. The algorithm uses adaptive distance norm to add its own norm induced matrix to each cluster. In the process of clustering, the best number of clusters is determined by introducing three indexes: partition coefficient (PC), classification entropy (CEC) and XB index. The experimental results show that the improved fuzzy C- means clustering algorithm can effectively mine the active hot areas of urban residents.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：C913.3

【参考文献】