摘要
针对共享单车停放点位置优化的问题,以现有的共享单车停放点位置作为初始聚类中心点,对共享单车的定位数据做聚类,并且提出了结合K-MEANS聚类和缓冲区分析的混合聚类方法,来解决初始聚类中心点相靠近时出现的共享单车错误分类问题。将最终聚类结果作为建议的共享单车停放点位置,进而依据建议的共享单车停放点位置判断现有的共享单车停放点位置的合理性,实验结果表明这种混合聚类方法的聚类效果理想。
Aiming at the problem of optimizing the location of shared bicycle parking spots,the existing shared parking spot position is used as the initial clustering center point,and the positioning data of shared bicycles are clustered,and a hybrid clustering method combining K-MEANS clustering and buffer analysis is proposed.The hybrid clustering method is to solve the problem of shared bicycle error classification when the initial cluster center points are close.The final clustering result is used as the recommended shared parking spot location,and then the rationality of the existing shared bicycle parking spot position is judged according to the recommended shared bicycle parking spot position.The experiment shows that the clustering effect of this hybrid clustering method is ideal.
作者
王霞
宋树华
汤军
刘远刚
WANG Xia;SONG Shu-hua;TANG Jun;LIU Yuan-gang(College of Geosciences,Yangtze University,Wuhan Hubei 430100;Jiangsu Zhitu Technology Co.,Ltd.,Yangzhou Jiangsu 225000)
出处
《数字技术与应用》
2019年第7期58-61,共4页
Digital Technology & Application
基金
国家自然科学基金青年基金资助(41701537)
关键词
混合聚类
共享单车
停放位置
合理性
hybrid cluster
sharing bicycles
parking location
rationality