需求和成本不确定的供应链综合生产计划鲁棒优化模型研究

发布时间：2018-02-27 00:38

本文关键词： 不确定性鲁棒优化综合生产计划 LINGO　出处：《南京理工大学》2014年硕士论文　论文类型：学位论文

【摘要】：在当今市场经济条件下,产品的生命周期越发短暂,市场竞争异常激烈,客户需求变化不定,加之供应链的网状结构等,这些因素使得供应链中存在诸多不确定性。供应链中的不确定性会给供应链带来一定的风险,严重时甚至摧毁整个供应链。因此需要运用不确定性优化方法来降低不确定性给供应链带来的风险,将不确定性因素降至最低,从而提高供应链中企业的竞争力。在生产管理中,综合生产计划是一项重要的技术水平规划；在规划领域,它是介于具有宽泛决策的长期规划和具有详细决策的短期规划之间的。因此,研究供应链的综合生产计划具有重要意义。本文研究的是这样一个供应链综合生产计划：供应链中有一个供应商,多个制造厂区和多个客户区。其中制造厂商所需的所有原材料都由这个供应商供应,而每个客户区所需的产品可以由任意一个或多个制造厂区供应。期间,每个制造厂区可以通过雇佣和解雇的方式调节劳动力,通过外包和加班的形式来改变生产力。本文首先对供应链综合生产计划这一概念进行界定,接着对供应链综合生产计划中存在的不确定性按照不同分类方式进行了归纳总结,在整个不确定性环境中,本文选取需求和成本这两种不确定性来进行研究,使用情景分析法对成本和需求不确定性进行描述,并选取鲁棒优化方法对供应链综合生产计划进行建模,分别为最小化损失和最大化市场需求满足率,并构造了合适的罚函数。该鲁棒优化模型不仅包含了损失最小化目标,还引进了市场满足率最大化这一目标。对于多目标求解,文中给出了处理方式,即采用归一化的手法,将多目标模型转化成单目标模型进行求解。为了验证模型的有效性和可行性,本文进行了算例设计,基于LINGO软件对所设计的算例求解。文章的最后对基于LINGO软件的求解结果进行了分析,不仅分析了目标函数随着风险系数变化的变化趋势,还分析了目标函数随权重系数变化的变化趋势。求解结果显示,本文构建的鲁棒优化模型具有很好的鲁棒性能。为了验证罚函数的重要性,文中还给出了不加罚函数的模型求解结果图形,并和加了罚函数的模型求解结果进行了对比。对比结果显示,加了罚函数的鲁棒优化模型鲁棒性能更好。最后本文给出了在目标函数1和目标函数2之间权衡的曲线图,决策者可以根据企业的实际情况,决定最小化损失和最大化市场满足率之间的权重。本文构建了综合生产计划的多目标鲁棒优化模型,通过算例求解,不仅验证了模型具有可行性和有效性,还给出了目标函数之间的权衡曲线图,对企业具有一定的参考价值。
[Abstract]:In today's market economy, the life cycle of products is more and more short, the market competition is extremely fierce, the customer demand is changeable, and the supply chain network structure, etc. These factors lead to a lot of uncertainties in the supply chain. The uncertainty in the supply chain will bring some risks to the supply chain. Therefore, it is necessary to use uncertainty optimization method to reduce the risk of the supply chain caused by uncertainty, and to minimize the uncertainty factors. So as to improve the competitiveness of enterprises in the supply chain. In production management, integrated production planning is an important technical level planning; in the field of planning, It is between the long-term planning with broad decision and the short-term planning with detailed decision. Therefore, it is of great significance to study the integrated production planning of supply chain. This paper studies such a supply chain integrated production plan: the supply chain has a supplier, multiple manufacturing areas and multiple customer areas, in which all the raw materials required by the manufacturer are supplied by this supplier. And the products needed for each customer area can be supplied by any one or more manufacturing areas. During this period, each manufacturing area can regulate the labor force by hiring and firing, Firstly, this paper defines the concept of supply chain integrated production plan, and then summarizes the uncertainty in supply chain integrated production plan according to different classification methods. In the whole uncertain environment, this paper chooses the uncertainty of demand and cost to study, and uses scenario analysis to describe the uncertainty of cost and demand. The robust optimization method is selected to model the integrated production planning of the supply chain, which is to minimize the loss and maximize the market demand satisfaction rate, respectively, and construct the appropriate penalty function. The robust optimization model not only contains the loss minimization objective. The goal of maximization of market satisfaction rate is also introduced. In order to verify the validity and feasibility of the model, a numerical example is designed. At the end of this paper, the solution results based on LINGO software are analyzed, not only the change trend of objective function with risk coefficient, but also the change trend of objective function with risk coefficient are analyzed. The change trend of objective function with weight coefficient is also analyzed. The results show that the robust optimization model constructed in this paper has good robustness. In order to verify the importance of penalty function, The result graph of the model without penalty function is given, and the result is compared with that of the model with penalty function. The robust optimization model with penalty function is more robust. Finally, the graph of tradeoff between objective function 1 and objective function 2 is given. Determine the weight between minimizing losses and maximizing market satisfaction. In this paper, a multi-objective robust optimization model of integrated production planning is constructed. The model is not only proved to be feasible and effective, but also the tradeoff curve between objective functions is given, which has some reference value for enterprises.
【学位授予单位】：南京理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP13;F274

【参考文献】