Design optimization of machines and manufacturing plants on the basis of component-oriented behavior models

机器和制造厂都是机电一体化系统。他们的设计需要几个学科的密切合作。机电一体化系统设计的目标是考虑预先指定的要求，在早期阶段创建、评估和优化不同的概念性解决方案。概念性的机电一体化解决方案是后续学科工作的基础。表现力设计模型可以减少实现时间和提高产品质量。该项目的目标是开发一种新的方法，以优化机械和制造工厂的机电行为，以满足技术和管理要求。基于数学优化方法，寻找出满足要求的机电一体化元件和尺寸参数。然后将竞争设计与最优参数进行比较，行为模型、需求模型和组件数据模型是优化的基础。为了进行行为建模，开发了一个包含机电、机械、液压和气动元件的库。在这个背景下，问题被解决了，应该如何设计模型，使其能够在早期阶段被参数化，但同时为优化提供可表达的结果。此外，还开发了一些方法来对需求进行建模，如成本、周期时间和能量消耗，并将其与行为模型相关联。所有参数和组件组合的潜在设计空间通常非常大。因此，零部件和尺寸参数的搜索需要一个健壮而高效的优化策略。一方面，这种优化策略充分结合了全局和局部求解器算法。另一方面，该策略包含了显著减少搜索空间的工程知识，所开发的模型用于建立可复用的库，并在原型中实现所述方法。模型和实现将在随后的几个测试用例中进行测试。

Machines and manufacturing plants are mechatronic systems. Their design requires the close cooperation of several disciplines. The aim of mechatronic system design is to create, evaluate and optimize different conceptual solutions at an early stage considering the beforehand specified requirements. The conceptual mechatronic solution is then the basis for the work in subsequent disciplines. Expressive design models allow to reduce realization-time and to increase product quality.The aim of the project is to develop a new methodology for optimizing the mechatronic behavior of machines and manufacturing plants with regards to technical and managerial requirements. On the basis of mathematical optimization methods those mechatronic components and dimensioning parameters are searched that optimally fulfill the requirements. The subsequent step is then to compare the competing designs with optimal parameterization.Models of behavior, requirements and component data are the basis optimization. For behavioral modeling a library with electromechanical, electrical, mechanical, hydraulic and pneumatic components is developed. In this context the question is addressed, how models should be designed that can be parameterized in the early phase, but at the same time deliver expressive results for optimization. In addition, methods are developed to model requirements, e.g. in terms of cost, cycle time and energy consumption, and to relate these to the behavior models.The potential design space of all parameter and component combinations is usually very large. Therefore, the search for components and dimensioning parameters requires a robust and efficient optimization strategy. On the one hand, this optimization strategy consists of an adequate combination of global and local solver algorithms. On the other hand, the strategy comprises engineering know-how to reduce the search space significantly.The developed models are used to build-up a reusable library and the methods are implemented in a prototype. The models and the implementation will be tested subsequently in several test cases.