论文部分内容阅读
本文通过对储货型生产企业的生产计划建立多目标规划模型,然后使用遗传算法来解决此类问题。在需求的预测上,本文通过Monte Carlo方法模拟得到需求量更具合理性,并应用此方法将利润、成本和机会损失的统计量作为适应值函数。本文还给出了不同个体多个目标的比较策略,力求在利润最大化,成本和机会损失最小化的前提下,为决策者提供多种不同策略的选择方案。这有利于决策者做出最有利于企业的决定。最后根据数值实验,可以看到边界初始化的效果要好于均匀初始化。
This paper establishes a multi-objective programming model for the production planning of storage-type manufacturing enterprises, and then uses genetic algorithm to solve such problems. In the forecast of demand, the Monte Carlo method is used to simulate the demand more rationally, and the statistic of profit, cost and opportunity loss is used as the fitness function. This article also gives a comparison strategy of multiple goals for different individuals, and tries to provide decision makers with a variety of options for different strategies under the premise of maximizing profit, minimizing cost and opportunity loss. This helps decision-makers make the most favorable decisions for the business. Finally, according to numerical experiments, we can see that the effect of boundary initialization is better than that of uniform initialization.