This paper presents a fuzzy genetic algorithm approach to generate, assess, and select a System of Systems (SoS) meta-architecture through coupled executable models. A type-1 fuzzy assessor is used to transform crisp performance attribute inputs into a meta-architecture assessment for use as part of the fitness function of a genetic algorithm. This algorithm is applied to the generation, assessment, and selection of a meta-architecture for a hypothetical lethal, non-line of sight fires SoS for which the key performance attributes are affordability, flexibility, performance, robustness, and reliability. Combinations of existing systems that have nonlinear interactions are assessed and compared to the United States Military Future Combat System. Results show that this approach produces architectures that provide the same performance without requiring the purchase of any new systems, potentially saving billions of dollars.
- Architecture,
- Evolutionary algorithms,
- Genetic algorithms,
- Purchasing,
- System of systems,
- Systems engineering,
- Architecture assessment,
- Executable model,
- Existing systems,
- Fitness functions,
- Future Combat Systems,
- Fuzzy - genetic algorithms,
- Non-line-of-sight,
- Nonlinear interactions,
- Computer architecture,
- Fuzzy-Genetic Algorithm
Available at: http://works.bepress.com/steven-corns/23/