Mining Multi-Objective Minimal Commitment Decision Significance via Cluster-and-Find-Changes
- View All
© 2017, American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved. In the design of aerospace systems, it is important to understand both the tradeoffs between desired objectives and the mapping between design decisions and objectives. In this work, we present a nonparametric analysis technique that combines clustering and matching to find “minimal change sets” which define a hierarchy of design commitment. In doing so, we hope to shift the thinking of clustering from looking for similar designs to improving understanding of the effects of design decisions on objectives. The technique is intended to be effective when the mapping from decisions to objectives is highly nonlinear and decisions interact. We demonstrate the technique on a simple continuous test problem, a simple discrete test problem, and demonstrate the technique’s application to a real-world model of a portfolio of guidance, navigation and control systems for space exploration.
author list (cited authors)