Visualization as a Decision Support Tool in Multidisciplinary Design

这笔赠款为开发数字和可视化工具提供资金，以帮助设计师在工业环境中做出决策，其中许多设计师正在互动开发相同的产品。 关键交付成果将是一个决策支持工具，设计人员可以查看本地和非本地最佳设计空间（即他们自己的最佳设计空间或另一个学科或子系统的最佳设计空间），预测和可视化本地和非本地决策的影响，并根据此信息确定良好和稳健的解决方案区域。 该工具的理论开发和实际使用的基础是图形变形（一种查看设计空间的可视化技术），博弈论（一组对决策者之间的交互进行建模的构造）和鲁棒设计技术（努力最大限度地减少不可控因素的影响，而不消除因素本身）。 如果成功的话，这项研究的结果将导致使用数学模型来近似子系统和设计师之间的接口的见解和改进。 这项工作的主要目标是开发一个强大的可视化方法，让多个设计师查看n维优化设计问题的图形。 这种能力将允许多个设计人员在同一产品上工作，以数字和图形的方式识别本地（例如，子系统或学科）或非本地（例如，其他学科或系统级）设计空间中的鲁棒、不敏感区域。 在这项工作中开发的决策支持方法和工具的主要优点将是提供信息的影响，由多个设计师对其他子系统的决策，并帮助设计师使用这些信息，使他们控制的变量和子系统的智能决策。

This grant provides funding for the development of a numerical and visualization tool to assist designers in making decisions in industrial environments wherein numerous designers are interacting to develop the same product. The key deliverable will be a decision support tool where designers can view local and non-local optimal design spaces (i.e. their own optimal design space or that of another discipline's or subsystem's), predict and visualize the effects of local and non-local decisions, and identify good and robust solution regions based on this information. The foundations for the theoretical development and practical use of the tool are Graph Morphing (a visualization technique to view design spaces), Game Theory (a set of constructs that model the interactions among decision makers), and robust design techniques (which strive to minimize the effects of uncontrollable factors without eliminating the factors themselves). If successful, the results of this research will lead to insights and improvements in using mathematical models to approximate the interfaces between subsystems and designers. The primary goal of this work is to develop a robust visualization approach to allow multiple designers to view n-dimensional optimal design problems graphically. This capability will allow multiple designers working on the same product to identify robust, insensitive regions in a local (for example, subsystem or disciplinary) or non-local (for example, other discipline or system level) design space, both numerically and graphically. The primary advantage of the decision support methods and tools developed in this work will be to provide information about the effects of the decisions being made by multiple designers on the other subsystems and to help designers use this information to make intelligent decisions about the variables and subsystems over which they have control.