Service Sites Selection for Shared Bicycles Based on the Location Data of Mobikes

Mobike is a leader in the bicycle-sharing industry, which commits to solving short-distance travel problems through the Internet. The rapid development of shared bicycles has been bringing very convenient for our travel. However, the problems of parking, high damage rates, and difficulty in reclaim have made many urban public space resources be taken up by these bicycles. In order to help the company maintain and manage its bicycles and service sites, the main contents of this paper focus on selecting the optimal service sites for Mobikes and planning the shortest circuit planning between service sites in each region. Our experimental data are the parking locations of Mobikes in Fuzhou city. The density-based clustering algorithm and an improved ant colony algorithm are adopted in this paper: 1) to divide a large data set into N small uncorrelate data sets, the density-based clustering algorithm reveals the high-density region by finding the low-density region of the dataset, and this type of algorithm is used for selecting the service sites and 2) the problem that starting from the warehouse center, then traversing all the service sites in the region, and finally returning to the warehouse center can be classified as an NP-hard problem. In this paper, an improved ant colony algorithm is used to obtain the optimal solution. The main purpose of finding the shortest circuit traversing all service sites in the region is to provide better support and maintenance for each service site.