摘要
研究并实现了一种结合核密度分析的DBSCAN空间聚类方法。首先对从马鞍山市出租车的轨迹数据中提取出上下客轨迹点数据进行预处理;再对上下客轨迹点数据进行聚类分析,识别居民出行热点区域;最后进行居民出行热点分析,并总结了居民出行的时空特征。
In this paper, we studied and implemented a DBSCAN spatial clustering method combining with kernel density analysis. Firstly, we extracted the trajectory point data of the upper and lower passenger from the taxi trajectory data in Ma'anshan City by pretreatment. Then, we clustered the trajectory point data of the upper and lower passenger, and identified the residential trip hotspot area. Finally, according to the hotspot analysis of resident trip, we summed up the spatio-temporal characteristics of the resident trip.
引文
[1]张朋东.基于浮动车数据的城市居民出行行为规律研究[D].长沙:中南大学,2012
[2]张治华.基于GPS轨迹的出行信息提取研究[D].上海:华东师范大学,2010
[3]Rosenblatt M.Remarks on Some Nonparametric Estimates of a Density Function[J].Selected Works in Probability and Statistics,2011,27(3):832-837
[4]Parzen E.On Estimation of a Probability Density Function and Mode[J].Annals of Mathematical Statistics,1962,33(3):1 065-1 076
[5]Borruso G.Network Density Estimation:a GIS Approach for Analyzing Point Patterns in a Network Space[J].Transactions in GIS,2010,12(3):377-402
[6]禹文豪,艾廷华,杨敏,等.利用核密度与空间自相关进行城市设施兴趣点分布热点探测[J].武汉大学学报(信息科学版),2016(2):221-227
[7]王明涛.移动对象轨道聚类算法的研究[D].南京:南京航空航天大学,2010