EAGER: A Control-Theoretic Approach for Designing Robust Complex Engineered Systems

这一探索性研究（EAGER）项目的早期概念拨款的目标是对适用于复杂工程系统设计的稳健设计的控制理论方法进行研究。鲁棒设计是在各种不确定性存在的情况下，具有相对不变性能的设计。目前的稳健设计方法主要适用于设计已建立的消费产品或更新现有的系统设计；这些方法不适用于大型复杂系统的设计，如航天器或智能电网，其中主要的不确定性，如环境，预算或新技术性能，是及时解决的。在这项工作中采用的研究方法将设计过程视为一个随机控制问题，其目标是在给定时间段内根据当前条件和可用信息最好地调整系统设计。设计过程用状态空间模型表示，用卡尔曼递归求解。研究任务包括制定一个激励航空航天设计问题，以说明复杂系统设计的挑战，收集环境数据以实现状态空间模型的求解，最后推导并求解随机控制方程。如果成功，本研究将为复杂系统稳健设计的一般理论提供基础。目前，对于复杂工程系统的稳健设计，缺乏理论和后续方法，导致经常出现成本和进度超支以及潜在的不良性能。本研究中采用的方法将是独特的，因为它将设计过程从在确定的设计阶段优化设计的静态过程转变为在整个生命周期中控制设计的动态过程。这项研究不仅有利于学术界，也有利于在政府机构和大公司工作的复杂工程系统的设计者。

The objective of this EArly-concept Grants for Exploratory Research (EAGER) Program award is to perform research into a control-theoretic approach to robust design, appropriate for complex engineered system design. A robust design is a design possessing relatively invariant performance in the presence of various uncertainties. Current approaches to robust design are suitable primarily for designing established consumer products or updating existing system designs; these approaches are not appropriate for the design of large complex systems, such as spacecraft or smart grids, where primary uncertainties, such as environmental, budgeting, or new technology performance, are resolved in time. The research approach taken in this work treats the design process as a stochastic control problem in which the goal is to best align the system design for the current conditions and information available in a given time period. The design process is represented using a state-space model which is solved using Kalman recursion. The research tasks include formulating a motivating aerospace design problem which illustrates the challenges of complex system design, collecting environmental data to enable solution of the state-space model, and finally deriving and solving the governing stochastic control equations.If successful, this research will provide the basis for a general theory of complex system robust design. Currently, there is a lack of theory, and subsequent methodology, for the robust design of complex engineered systems, leading to frequent cost and schedule overruns and potentially undesirable performance. The approach taken in this research will be unique in that it will transform the design process from a static process of optimizing the design in a defined design phase to a dynamic process of controlling the design throughout its lifecycle. The proposed research will benefit not only the academic community, but also designers of complex engineered systems working in government agencies and large corporations.