DEVELOPMENT OF EVOLUTIONARY ALGORITHMS FOR FACILITY ARRANGEMENT PROBLEMS CONSIDERING CHANGES WITH MULTIPLE PERIODS

对于考虑多周期变化的设施布局问题，应对每个周期的设施布局计划进行规划，而不是只规划一个周期的规划。通过建模，在考虑产品需求变化的成本约束下，确定各设施的运行/终止以及在终止设施中的生产整合，以最大化设施的生产能力。通过数值实验研究了遗传算法、模拟退火法和禁忌搜索法的应用，验证了遗传算法的有效性。设施超过20个，成本降低率低于0.8的问题难以解决。对于产生式集成的求解方法，利用启发式局部搜索提高了搜索效率。

For facility layout problems considering changes with multiple periods, facility layout plans in every periods should be planned not only a plan in one period. In this research, the operation/termination of each facility and integration of production in terminated facilities are decided under the cost constraint considering changes of demands of the product to maximize the production capacity of facilities with modeling. In this research, application of genetic algorithm (GA), simulated annealing and tabu search method are considered with numerical experiments and the effectiveness of GA is verified. It is difficult to solve problems of cases of more 20 facilities and ratio of cost reduction is less than 0.8. For solving method of integration of production, searching effectiveness is improved by using heuristic local search.