Systematic Mathematical Strategies for Stochastic Modeling and Uncertainty in Production Planning and Scheduling

摘要：Marianthi Ierapetritou机构：罗格斯大学提案编号：0625515标题：随机建模和不确定性的系统数学策略在生产计划和生产计划中项目摘要单个生产设施的过程操作的优化涉及大量需要考虑的相互依赖的因素，涉及经济最优性，需求满足，容量限制，生产配方，顺序相关任务，现有的模型试图在问题的现实描述和解决方案程序方面的可用能力之间进行折衷。该项目的目标是处理生产规划中的不确定性的方法。计划的综合框架针对优化生产系统领域的一个关键问题，即不确定性问题。主要结果是提供替代解决方案的决策者，表现出不同的特点，在面对不确定性，即预期的盈利能力，灵活性和鲁棒性。为了实现这一目标，将根据有关不确定性的信息水平研究以下想法：不确定性空间中的敏感性分析和参数规划。新的想法将进行调查，这将使有效的生成参数的解决方案的情况下，没有足够的信息来描述不确定性。 在早期决策阶段考虑风险和盈利能力的多目标优化，将产生适当平衡不同目标的解决方案。这种方法将采用的情况下，有可用的信息来描述不确定的参数。其最终目标是在考虑不确定性问题的情况下，优化大规模单设施生产工厂的生产计划操作。来自石化公司、快速消费品制造和化工公司的实际案例研究和实际例子将被用来测试和验证所提出的分析工具的结果。广泛的影响工作的影响超出了具体的工业部门的运作，因为一般的想法和调查结果可以适用于许多类型的生产设施。通过建立一个网络环境传播结果，将能够更好地与工业界互动。该项目的目标之一是在化学工程课程中整合过程操作、结果和案例研究的基本概念。一个重组的设计课程和一个新推出的跨学科系统工程课程将被用作整合工具。

ABSTRACTPI: Marianthi Ierapetritou Institution: Rutgers UniversityProposal Number: 0625515Title: Systematic Mathematical Strategies for Stochastic Modeling and Uncertainty in Production Planning and SchedulingProject SummaryOptimization of process operations for a single production facility involves a large number of interdependent factors that need to be considered involving economic optimality, demand satisfaction, capacity limitations, production recipe, sequence dependent tasks, etc. Existing models attempt to develop a compromise between a realistic description of the problem, and available capabilities in terms of solution procedures. This project targets methodologies to deal with uncertainty in production planning. The integrated framework that is planned targets a key issue in the area of optimizing production systems, which is the problem of uncertainty. The main outcome is to provide alternative solutions to the decision maker that exhibit different characteristics in the face of uncertainty, namely expected profitability, flexibility, and robustness. To achieve this target, the following ideas will be investigated depending on the level of information available about uncertainty: Sensitivity analysis and parametric programming in the uncertainty space. New ideas will be investigated that will enable the efficient generation of parametric solutions for the case where there is not sufficient information to characterize uncertainty. Multi-objective optimization considering risk and profitability in the early decision stages that will result in solutions that balance appropriately the different objectives. This approach will be employed in the cases where there is available information to describe the uncertain parameters. The ultimate goal is to optimize the production planning operations in large-scale, single-facility production plants considering the issues of uncertainty. Realistic case studies and actual examples from petrochemical companies, fast moving consumer goods manufacturing and chemical companies will be utilized to test and verify the results of the proposed analysis tools. Broad Impact The impact of the work goes beyond specific industrial sector operations since the general ideas and findings can be applied to many types of production facilities. The dissemination of the results through the development of a web-based environment will enable better interactions with industry. Among the targets of this project is the integration of the basic concepts of process operations, results and case studies within the chemical engineering curriculum. A restructured design course and a newly introduced interdisciplinary systems engineering course will be used as integration tools.