Premium
Utilizing a maximin optimization approach to maximize system resiliency
Author(s) -
Liaghati Hassan Robert,
Mazzuchi Thomas A.,
Sarkani Shahram
Publication year - 2021
Publication title -
systems engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.474
H-Index - 50
eISSN - 1520-6858
pISSN - 1098-1241
DOI - 10.1002/sys.21570
Subject(s) - minimax , resilience (materials science) , process (computing) , genetic algorithm , computer science , optimization problem , mathematical optimization , mathematics , physics , thermodynamics , operating system
Abstract Complex systems typically have a large parameter space and by definition have the potential for strong emergent behavior, especially when operating in a complex environment. A method is needed to design a system that will be resilient in a complex target operating environment. Here, we propose a maximin optimization approach to support the exploration of system design, based on co‐evolutionary system/environment models, which emphasizes the importance of the operating environment in the system evolution, resulting in increased system resilience. In this paper, this method is demonstrated on an air defense system and shows an increase in resilience when compared to the system resulting from the application of a genetic algorithm such as has been previously demonstrated for other systems. For the air defense system used here, the proposed maximin optimization process significantly ( p < . 001 , via a one‐sided paired t ‐test) increased the resulting system resilience from 0.64 (the resilience of the system designed by a genetic algorithm) to 0.92 (the resilience of the system designed by maximin optimization).