摘要
丰富的居民出行行为信息对挖掘城市热点区域以及居民出行模式有很大的帮助,并且对更好地满足居民出行需求也有一定的启示作用.最新的相关研究主要聚焦于城市中区域之间的空间移动模式,但并不能识别移动模式发生的时间以及持续的时长.针对这一问题,提出具有时空特性的区域移动模式挖掘算法STMPZ(Spatio-Temporal based Movement Patterns between Zones).该算法在DBSCAN(Density-based Spatial Clustering of Applications with Noise)算法的基础上,通过将对象从点扩展成一条出行OD(Origin-Destination)记录,并引入时间特性,最终可以挖掘出具有时空特性的区域移动模式.为了验证所提出算法的可行性和有效性,利用真实的上海地铁通勤数据集进行实验,实验结果表明,该算法可以快速有效地检测出具有高覆盖率和准确率的区域移动模式.此外,该算法也可以通过修改聚类过程的参数应用于其他区域或类型的交通数据.
The extensive information of residents'travel behavior is very helpful for mining hotspots in cities and the patterns of residents'travel.It also has a certain enlightening effect on better meeting residents'travel needs.The recent related researches mainly focus on the patterns of spatial movement among regions in cities,which cannot identify the occurrence time and duration of movement patterns.Based on the existing researches and problems,a new algorithm STMPZ (Spatio-Temporal based Movement Patterns between Zones)is presented for mining movement patterns between zones with spatiotemporal characteristics.Based on the DBSCAN(Density-Based Spatial Clustering of Applications with Noise),the algorithm extends the point object to a trip record,and introduces time characteristic which can eventually mine the regional moving patterns with time and space characteristics.Because the STMPZ algorithm considers the spatial proximity and temporal proximity between objects in the process of regional mobile pattern mining,it does not only need to define the time interval,but also compensates for the traditional algorithm,even through the predefined time intervai.It is also impossible to recognize the defect of the area movement pattern which occurs at any time and lasts for an arbitrary length of time.In order to illustrate the feasibility and effectiveness of the proposed algorithm,real subway data sets in Shanghai were used for experimentation.The results show that the algorithm can effectively mine movement patterns between areas with high coverage and accuracy.The algorithm can also be applied to other areas or types of traffic data by modifying the parameters of the clustering process.
作者
周星星
张海平
吉根林
Zhou Xingxing;Zhang Haiping;Ji Genlin(School of Computer Science and Technology,Naniing Normal University,Nanjing,210023,China;School of Geographic Science,Nanjing Normal University,Nanjing,210023,China;Jiangsu Center for Collaborative Innovation in Geographical Information Resource Development and Application,Nanjing,210023,China)
出处
《南京大学学报(自然科学版)》
CAS
CSCD
北大核心
2018年第6期1171-1182,共12页
Journal of Nanjing University(Natural Science)
基金
国家自然科学基金(41471371)
关键词
区域移动模式
时空分析
出行行为
移动模式挖掘
movement pattern between areas
spatio-temporal analysis
travel behavior
movement pattern mining