Using genetic algorithms for naval subsystem damage assessment and design improvements

Some auxiliary systems of next generation naval ships will utilize distributed automatic control. Such distributed control systems will use interconnected sensors, actuators, controllers and networking components to diagnose and reconfigure the auxiliary systems. Testing these systems will be difficult with traditional methods of fault analysis due to the interconnected and automatic nature of these subsystems. We have designed a suite of genetic algorithms to find interesting and hidden damage scenarios in a testbed of a naval subsystem. Given this knowledge, we use a genetic algorithm to improve upon the design of this subsystem. Categories and Subject Descriptors

• Type: Conference
• Year: 2007
• Where: GECCO
• Authors: Christopher McCubbin, David Scheidt, Oliver Bandte, Steven Marshall, Iavor Trifonov
• Keywords: Auxiliary Systems, GECCO 2007, Genetic Algorithm, Such Distributed Control