Dataset

Step 1: We begin by defining our fundamental geographical unit of analysis, such as the population centroid, and then create aggregated demand zones, with information on overall demand provided at each zone.
We divided the city of Delhi into 1KM x 1KM square grids using ArcGIS software and plotted the centroid of each grid, obtaining the latitude and longitude of each centroid location. These centroids serve as customer zones that collection centers must service.
Step 2: Search for potential collection sites.
We consider gasoline pumps/gas stations, Residential Welfare Association offices in Delhi, and vehicle dealer locations in Delhi as potential collection center sites for our problem. Gas stations, vehicle dealerships, and gasoline pumps were located using Google Maps, and the addresses of residential societies were obtained from the Delhi government's official website.
Step 3: Create a network-based demand zone for each demand node.
The network demand area determines the distance customers are willing to travel to return their used batteries. We know the locations of all demand and potential collection sites in our scenario. Using Microsoft Bing APIs, we calculated the distance between all demand locations and collection sites.
Step 4: Uncover the most suitable locations for facilities like collection centers within each demand region and identify the optimal incentives to be offered through various collection channels to encourage customers to return the used batteries.

第一步，我们首先定义分析的基本地理单元（如人口质心（population centroid）），随后构建聚合需求区，每个区域均包含整体需求信息。我们利用ArcGIS软件将德里市划分为1千米×1千米的方形网格，绘制每个网格的质心并获取其经纬度坐标。这些质心构成了回收中心需服务的客户区域。
第二步，筛选潜在回收站点。我们将德里市的加油站（gasoline pumps/gas stations）、居民福利协会（Residential Welfare Association）办公室及汽车经销商网点列为潜在回收中心选址。加油站、汽车经销商网点通过谷歌地图（Google Maps）定位，而居民社区地址则取自德里市政府官方网站。
第三步，为每个需求节点构建基于网络的需求区。该网络需求区定义了客户为归还废旧电池愿意出行的最大距离。我们已知场景中所有需求点及潜在回收站点的位置，通过微软必应应用程序接口（Microsoft Bing APIs）计算了所有需求点与回收站点之间的距离。
第四步，在各需求区内确定回收中心等设施的最优选址，并明确通过多种回收渠道提供的激励措施，以鼓励客户归还废旧电池。