一种基于市场真实情况的ATM选址模型的研究及应用

摘要

The deployment of ATMs is influenced by many factors, such as population density, business, peer competition,and deployed ATMs.Aiming at the problem of ATM location, a solution algorithm based on Voronoi diagram and K-mediods clustering is proposed and designed,and an ATM location model which is more in line with the market real situation is constructed.The proposed ATM location model has been successfully applied to a bank in Shenzhen and reduced the weighted distance by 11.8% from users to the bank's ATM deployment site by selecting only four new ATM deployment sites,which achieved a reasonable application effect.%自动取款机(ATM)的部署受到人口密度、商业、同行竞争和已部署机具场景等诸多因素影响.针对ATM选址问题,提出和设计一种基于Voronoi图和修改的K中心点聚类的求解算法,构建更符合市场真实情况的ATM选址模型.该ATM选址模型已成功应用于深圳市某银行,仅选取了4个新ATM的部署地点,使用户到该银行ATM部署点的加权距离减小了11.8%,取得了合理应用效果.