Resilience-Based Optimization of Space Systems

With future missions of increasing complexity, duration, distance, and uncertainty, there has been a growing need for methods and tools that can permit the effective formation of early stage conceptual designs that are not only cost-effective, but also productive and resilient to failures. Work performed during Phase I, involved development of a method and tools to address these challenges for space-based systems. However, further development is needed to expand these approaches, making them viable for a more general set of problems. Furthermore, improvements both to the quality of solutions obtained, and the efficiency with which these solutions are obtained is necessary. Finally, development of flight software to improve performance of complex space systems will demonstrate real-world benefits of this Phase II effort.This proposal addresses these issues with the following innovations:Development of a method and a set of tools that evaluate and optimize any type of arbitrary systemImproved metrics for evaluating and optimizing systems with multiple objectivesDevelop a multi-threaded algorithm that can benefit from parallel computingDevelopment of an efficient method to quantify and propagate uncertainty associated with the systemDevelopment of a general set of tools to capture, archive, and analyze data flows on complex, message-oriented space-borne data busesThe significance of the innovations is that the proposed methods and tools will:Permit the optimization of any arbitrary type of system or system-of-systemsProvide higher quality solutions to problems seeking to optimize multiple objectives simultaneously in an efficient mannerAllow for the efficient quantification and propagation of uncertainty associated with the systemAllow for the inclusion of onboard data-flow considerations earlier in the system-level design phase of complex space systems as a means of increasing the robustness of the design to growth in data volumes or increases in data rates