Novel Modeling and Analytical Approaches for Primary Pharmaceutical Manufacturing Scheduling and Related Problems

【摘要】针对初级药品生产调度及相关问题的新型建模和分析方法[j]: Subhash C. Sarin和Hanif D. sherali该基金为开发优化模型和算法提供资金，用于分析药品供应链的初级生产操作阶段，该阶段负责活性药物成分的生产。所设计的方法将确定药品的最佳分配和排序，以便以最小的生产成本（包括所需设置和库存的贡献，以实现快速周转）满足客户的要求，给定一组具有指定容量的设备以及每种产品的产品组合和需求率。将开发一种基于分解的方法来确定（接近）最优的批处理、分配和排序决策。在实践中出现的一些重要的特殊情况也将进行研究。所开发的模型和解决方案方法将在勃林格殷格翰化学工厂遇到的实际问题实例中进行测试，该工厂位于弗吉尼亚州彼得堡，致力于为制药行业生产活性药物成分和中间体。如果成功，本研究的结果将通过减少实践中的周期时间延迟来提高对市场需求的响应能力，同时降低制造成本。所研究的模型包括作为子结构的高多重不对称旅行商问题和经典的作业车间调度问题。基于所建议的研究工作的多面体分析也将有助于加强这些问题的独立解决，这些问题本身就很重要。此外，拟议的研究将有助于为分析其他批量生产设施（如半导体制造）提供见解和计算工具。

AbstractNovel Modeling and Analytical Approaches for Primary Pharmaceutical Manufacturing Scheduling and Related ProblemsPIs: Subhash C. Sarin and Hanif D. SheraliThis grant provides funding for developing optimization models and algorithms for analyzing the primary manufacturing operational stage of the pharmaceutical supply chain, which is responsible for the production of active pharmaceutical ingredients. The designed approach will determine optimal assignment and sequencing of pharmaceutical products to various processing bays of a primary manufacturing facility in order to meet customer requirements at minimal production costs (which include contributions from both the setups required and inventory incurred to achieve a quick turn-around), given a set of equipments with specified capacities as well as a product-mix and demand rate for each product. A decomposition-based methodology will be developed for determining (near-) optimal batching, assignment, and sequencing decisions. Certain important special cases that arise in practice will also be studied. The models and solution methodologies developed will be tested on real-life problem instances encountered by the Boehringer Ingelheim Chemical plant, located in Petersburg, Virginia, which is devoted to the production of active pharmaceutical ingredients and intermediaries for the pharmaceutical industry.If successful, the results of this research will lead to increasing the responsiveness to market demand by reducing cycle time delays in practice, while curtailing manufacturing costs. The models to be investigated include as substructures the high multiplicity asymmetric travelling salesman problem and the classic job shop scheduling problem. The polyhedral analysis based on the proposed research effort will also serve to enhance the independent solution of these problems, which are important in their own right. In addition, the proposed research will contribute toward providing insights and computational tools for analyzing other batch production facilities, such as in semiconductor manufacturing.