多阶段零售店选址的模型与算法研究

Model and algorithm of multistage facility location for retail trade

针对零售店的选址问题,设计了分段效用函数,改进了现有竞争性设施选址模型中效用函数的设计不足,把消费者居住地到零售店的距离对效用的作用分成无影响、有直接影响和绝对影响三种情况考虑。在此基础上,建立了多阶段零售店选址模型,决策者对不同阶段的零售店重新布局,使其与消费者需求分布和零售店经营成本变动相匹配。根据模型特点设计了分散搜索算法,分别用Lingo软件和分散搜索算法编程对一组算例进行计算比较,两种算法的运算结果显示,分散搜索算法的运行速度快而且收敛性好。

Aiming at the problem of retail site location, to improve existing utility functions in competitive facility location models, a piecewise utility function was designed, in which the influence of distance from the inhabitancy of customers to retail sites was discriminated by three forms in utility function： the distance without affecting customers＇ utility, directly affecting customers＇ utility and absolutely affecting customers＇ utility. Under this cir- cumstance, a multistage retail sites location model was built so that matching customers＇ demand distribution and running cost of retail sites at different stages would be matched, and an effective scatter search heuristic algorithm was devised according to the model characteristics. An example and a set of simulation experiments were calculated by the proposed scatter search heuristic algorithm and Lingo system. Solution results showed that the precision of solutions and solving times with scatter search algorithm were superior to those of Lingo system.