摘要
为合理进行共享汽车网点选址,提高共享汽车使用率,以成都市滴滴快车平台出行数据为基础,将K-means算法与遗传算法(Genetic Algorithm,GA)进行改进,提出了一种新型算法——KM-GA算法,进行共享汽车网点选址研究。针对成都市青羊区用户出行需求得到15个共享汽车网点,并将该算法得出的结果与常用选址聚类算法Kmeans、DBSCAN以及支持向量机(SVM)进行对比。结果表明,无论是网点距离和、需求量、覆盖率还是土地价格,KM-GA算法得到的结果均优于其他3种算法。该算法合理地确定了K值的取值,避免了传统K-means算法易陷入局部最优解的弊端,为共享汽车选址提供了新的思路。
In order to rationally select the location of shared cars outlets and increase the utilization rate of car-sharing,this paper used KM-GA algorithm which combines K-means algorithm with Genetic Algorithm(GA)to study site selection by using the trip data of Chengdu Didi Express platform.According to the travel needs of users in Qingyang District,Chengdu,15 car-sharing outlets were derived,and the results obtained by this algorithm were compared with commonly used location clustering algorithms K-means,DBSCAN and support vector machine(SVM).The experimental results show that the results obtained by the KM-GA algorithm are better than the other three algorithms,whether it is the distance and the demand,the coverage rate or the land price.This algorithm reasonably determines the value of K and avoids the disadvantage of the traditional K-means algorithm that is easy to fall into the local optimal solution.It provides new ideas for the location of shared cars.
作者
王玥
李文权
梁爽
余静财
WANG Yue;LI Wen-quan;LIANG Shuang;YU Jing-cai(School of Transportation,Southeast University,Nanjing 210096,China)
出处
《武汉理工大学学报》
CAS
2021年第2期79-85,共7页
Journal of Wuhan University of Technology
基金
国家重点研发计划(2018YFB1601001)
江苏省研究生科研与实践创新计划(SJCX20_0046)
中央高校基本科研业务费专项资金资助项目(3221002121D)
