Methodology for multiple response optimization of lean and resource efficient manufacturing systems

对于制造企业来说，除了经济优化之外，有针对性地有效利用现有资源也变得越来越重要。基于精益理念的制造系统的经济和组织优化主要与资源节约型制造的目标相适应。然而，这两项原则的兼容性是有限的。这就需要对将要改进的制造系统进行单独的分析。此外，必须考虑潜在改进措施的利弊。本提案旨在开发一种基于仿真的方法，用于精益和资源高效制造系统的多目标优化，并将理想的典型制造系统的相互依赖关系推广到精益和资源高效制造系统的指导方针中。为了预测理想的典型制造系统的各种配置的影响，离散事件模拟的材料、能量和信息流将与来自精益和资源高效制造目标标准的评估模型相结合。通过实验设计，将高斯过程模型拟合到复杂仿真模型的行为中。这些元模型允许在短时间内计算系统行为的近似，因此，制造系统的影响和相互依赖性的广泛多目标优化和统计分析将成为可能。由于计算时间长，模拟模型几乎不可能做到这一点。所分析的精益和资源高效制造系统的影响和相互依赖性将被转化为每个理想的典型制造系统的指导方针。随后，将通过选定的工业案例研究对方法和拟定的准则进行检验，并可能加以调整。最后，利用所分析的工业制造系统的元模型确定其最优目标状态。在此基础上确定实际的实践措施，选择最合适的措施组合，确定最优的实施顺序。

For manufacturing companies, the targeted and efficient use of available resources is gaining in importance, alongside economic optimization. The economic and organizational optimization of manufacturing systems according to the lean philosophy is mainly compatible with the goals of resource-efficient manufacturing. However, there are limits to the compatibility of the two principles. This necessitates individual analyses of to-be improved manufacturing systems. Furthermore, advantages and disadvantages of potential improvement measures must be considered. This proposal aims at developing a simulation based methodology for the multi-objective optimization of lean and resource-efficient manufacturing systems and for the generalization of ideal typical manufacturing systems' interdependencies into guidelines for lean and resource efficient manufacturing systems. To predict the effects of various configurations of ideal typical manufacturing systems, the material, energy, and information flows of a discrete event simulation will be coupled with an assessment model from the target criteria of lean and resource-efficient manufacturing. Using design of experiments, Gaussian process models will be fitted to the behavior of the complex simulation models. These meta-models allow the approximation of the system behavior to be computed in a short period of time and, thus, extensive multi-objective optimization and statistical analysis of a manufacturing system's effects and interdependencies will be enabled. This would hardly be possible with the simulation model, due to long computational time. The analyzed effects and interdependencies of lean and resource-efficient manufacturing systems will be translated into guidelines for each ideal typical manufacturing system. Subsequently, the methodology and the developed guidelines will be tested and possibly adapted by selected industrial case studies. Finally, the optimal target states of the analyzed industrial manufacturing systems will be identified using their meta-models. Based on this, real practice measures will be determined, the most suitable combination of measures will be selected, and their optimal implementation sequence will be determined.