Holistic Design Optimization in Mechatronics

Design of modern mechatronic systems can be an intimidating task. The underlying problem lies in that several different engineering domains merge into one product, creating integration issues. Commonly used design methodologies are based on optimizing the different domains separately; hence creating a suboptimal final system. This paper contributes to the field by describing and discussing a holistic approach to design and optimize mechatronic products, especially useful in an early design phase. Specifically, this paper extends previously published work by taking control aspects into account, as well as enabling the use of multiple optimization criteria. The design approach described is based on using simplified, static models, to dimension and describe physical properties of structure and transducer components, while transfer function models are used for control design and behavioral modeling. To enable time-efficient optimization, computationally inexpensive ways to evaluate structural and behavioral properties are sought. An evolutionary optimization algorithm is used to evaluate the component models and by doing that derive an optimal solution.