Modular manufacturing is gaining popularity in a variety of industrial applications due to potential cost-savings as well as process flexibility that can be achieved with the use of small and standardized modules. In this work, a framework for supply chain optimization is proposed that ensures production feasibility with the help of historical process data for individual process modules and machine learning-based feasibility analysis. A supply chain optimization problem is formulated where the binary variables represent facility locations, and the integer variables correspond to the number of modular equipment installed. Results demonstrate that the tradeoff between centralized and distributed manufacturing and the effect of economies of numbers on the cost of the supply chain can be studied by solving the problem.

由于潜在的成本节省以及使用小型和标准化模块可以实现的工艺灵活性，模块化制造在各种工业应用中变得越来越流行。在这项工作中，提出了一个用于供应链优化的框架，该框架可借助各个过程模块的历史过程数据和基于机器学习的可行性分析来确保生产的可行性。提出了一个供应链优化问题，其中二进制变量代表设施位置，整数变量对应于已安装的模块化设备的数量。结果表明，可以通过解决该问题来研究集中制造和分布式制造之间的权衡以及数字经济对供应链成本的影响。