Abstract: Bicycle sharing has not only enriched the travel options of residents, but also saved time, money and fitness, and has become an important means of transportation for people traveling short distances. However, the number of faulty bicycles caused by normal wear and tear and human damage in the use of shared bicycles is very large, and the task of recycling and repairing or scrapping them is very heavy, which has become a problem of reverse logistics in the shared bicycle market.
The current academic research on the logistics of the shared bicycle market focuses more on the problems of placement scheduling and distribution of stationing points. In terms of the recycling logistics of shared bicycles, the existing research mainly focuses on the route planning in the recycling process of faulty bicycles. This paper uses the Wuhan shared bicycle market as an example to explore the optimal location strategy of recycling centers based on the clustering analysis of faulty shared bicycle scrapping points and the center of gravity method based on transportation cost orientation, which helps to solve the problem of choosing the location of shared bicycle recycling centers.
This paper argues that the optimization of the location of faulty shared bicycle recycling centers can help reduce the operating costs of faulty bicycle recycling, improve the efficiency of faulty bicycle recycling, and promote the development of reverse logistics in the shared bicycle market. The idea of optimizing the location of faulty bicycle recycling centers is: (1)Faulty bicycle scrapping and overhaul. After a shared bicycle is put into a designated area, it needs to be scrapped and overhauled under two circumstances: Users mark it as a faulty bicycle through the APP, or the regional manager regularly checks the GPS information of the backstage vehicles and the condition of the vehicles on a daily basis, and marks the backstage of faulty bicycles that have safety hazards or need to be suspended. (2)Manual identification of faulty bicycles. The operation background sends the acquired information to the maintenance person in charge of the designated area, screens out the vehicles that need to be recycled and overhauled, and carries out secondary marking to facilitate subsequent recycling processing. (3)Clustering of scrapping points. The initial location of the faulty vehicle is marked as scrap point by the background, and according to the result of data processing, the faulty vehicle is clustered by K-means algorithm for the area to which the recycling center belongs, and according to the coordinates of the scrap point clustering, the daily arrangement of transportation vehicles and manuals will recycle the vehicles at the scrap point to the scrap center according to the established recycling route. (4)Execution of recycling tasks. The faulty bikes are transferred from the scrap center to the recycling center for overhaul and then put back, and the recycling task is completed.
In this paper, we first construct a clustering model for the scrapping centers of faulty bikes, use the “elbow” method to determine the best cluster number, standardize the elemental data, determine the k initial cluster centers according to the best cluster number, and assign the samples to the nearest cluster according to the principle of the shortest Euclidean distance. The mean value of samples in each cluster is used as the new clustering center, and the above steps are repeated until the clustering center no longer changes, which finally makes all samples form the best clustering result, thus completing the clustering of scrap centers. The center of gravity method is based on the principle of optimal cost, and the recycling center is equivalent to the clustering object in the k-means clustering algorithm, and the center of gravity of the system is the best setting point for the recycling center of the faulty shared bicycle.
The simulation validation of the model shows that the optimized location points of the recycling center not only reduce the cost of faulty shared bicycle recycling, but also have lower overall operation cost and higher efficiency of faulty shared bicycle recycling compared with the three existing recycling centers in Wuhan city which are distributed farther away, and facilitate the sub-regional operation and management of shared bicycles. It is proved that the K-means clustering algorithm and the center of gravity method are not only simple and feasible, but also convenient and fast, and this model can take into account many factors compared with the realistic site selection method that only considers cost. The K-means clustering algorithm and the center of gravity method are used to determine the location of the recycling center, which is suitable for all areas of the city.

Key words: shared bicycle, recycling center, site selection optimization

摘要： 共享单车已日渐成为人们短途出行的重要交通工具,但共享单车市场一贯采用“重投放,轻维护”的发展模式,而共享单车使用中的正常损耗及人为破坏等造成的故障车数量却十分庞大,对其回收修复或报废的任务繁重,这也成了共享单车市场逆向物流亟待解决的难题。该文立足武汉共享单车市场,通过对故障共享单车报废点的聚类分析,基于运输成本导向,使用重心法探寻回收中心最佳选址点,以解决故障共享单车回收成本和效率问题。对模型的模拟验证表明,优化后的回收中心选址点不仅能降低故障共享单车回收成本,而且相比武汉市现有的三个分布较远的回收中心,其总体运营成本更低,故障共享单车回收效率更高,便于共享单车的分区域运营管理。事实证明,基于K-means聚类算法与重心法确定回收中心选址问题不仅操作简单,可行性高,而且方便快捷,相较于现实中单一考虑成本等的选址方式,此模型更能兼顾多方面因素,优势明显。基于K-means聚类算法与重心法来确定回收中心选址,适用于城市的各个区域,选点精确又方便高效,模型具有较强的实用性。

关键词: 共享单车, 回收中心, 选址优化