模拟植物生长算法在设施选址问题中的应用

Application of plant growth simulation algorithm on solving facility location problem

以模拟植物生长算法为工具,提出了一种解决设施选址问题的智能优化算法.结合配送中心选址的实际案例,将模拟植物生长算法与遗传算法的计算结果进行比较,结果表明该算法比遗传算法在精度上有所提高;在此基础上,以50个随机选取的用户为背景,解决了韦伯型多设施选址问题.不同于其它启发式算法,模拟植物生长算法在得到全局最优解的同时,还可以根据设施数量的不同,将全局最优解与局部最优解进行组合,可以建立整体最优的设施布局.本算法在应用中显示了精确性、稳定性和通用性特点,是模拟植物生长算法在解决选址问题上的具体应用.

Based on Plant Growth Simulation Algorithm （PGSA）, we propose a intelligence optimization algorithm for solving facihty location problems. We compare the calculating results of PGSA with Genetic Algorithm （GA） for distribution center location problem, and the result approves PGSA is better than GA on accuracy. Further more, selecting 50 customers randomly, we solve Weber multi-facility location problem. Differed from other heuristic algorithms, PGSA can find global optimal solutions. Meanwhile, according to the different facility numbers, we combine global and local optimal solutions, set up optimal facility location arrangement as a whole. The algorithm herein shows its accuracy, astringency and generalization. It is an actual application of PGSA on solving location problems.