Designing mechatronic systems is known to be a very complex and tedious process due to the high number of system components, their multi-physical aspects, the couplings between the different domains involved in the product, and the interacting design objectives. This inherent complexity calls for the crucial need of a systematic and multi-objective design thinking methodology to replace the often-used sequential design approach that tends to deal with the different domains and their corresponding design objectives separately leading to functional but not necessarily optimal designs. Thus, a new approach based on a multi-criteria profile for mechatronic systems is presented in this paper for the conceptual design stage. Additionally, to facilitate fitting the intuitive requirements for decision-making in the presence of interacting criteria, three different methods are proposed and compared using a case study of designing a vision-guided quadrotor drone system. These methods benefit from three different aggregation techniques such as Choquet integral, Sugeno integral and fuzzy-based neural network. To validate the decision yielded by the results of global concept score for each aggregation methods, a computer simulation of a visual servoing system on all design alternatives for quadrotor drone has been performed. It is shown that although the Sugeno fuzzy can be a useful aggregation function for decisions under uncertainty, but the approaches using Choquet fuzzy and fuzzy integral-based neural network seem to be more precise and reliable in a multi-criteria design problem where interaction between the objectives cannot be overlooked.

中文翻译：

---

机电系统设计中概念评估的多准则模糊决策支持：四旋翼设计案例研究

众所周知，设计机电一体化系统是一个非常复杂和乏味的过程，因为系统组件数量多、物理特性多、产品涉及的不同领域之间存在耦合以及交互设计目标。这种固有的复杂性迫切需要一种系统的、多目标的设计思维方法来取代经常使用的顺序设计方法，这种方法倾向于分别处理不同的领域及其相应的设计目标，从而实现功能性但不一定是最佳设计。因此，本文在概念设计阶段提出了一种基于机电系统多标准配置文件的新方法。此外，为了便于在存在交互标准的情况下满足决策的直观要求，使用设计视觉引导四旋翼无人机系统的案例研究，提出并比较了三种不同的方法。这些方法受益于三种不同的聚合技术，例如 Choquet 积分、Sugeno 积分和基于模糊的神经网络。为了验证每种聚合方法的全局概念评分结果所产生的决策，对四旋翼无人机的所有设计备选方案进行了视觉伺服系统的计算机模拟。结果表明，虽然 Sugeno 模糊对于不确定性下的决策可以是一个有用的聚合函数，但是使用 Choquet 模糊和基于模糊积分的神经网络的方法在多准则设计问题中似乎更加精确和可靠，其中目标不容忽视。