Network-centric fuzzy resource manager: structure and validation
James F. Smith, Zachary Firth · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2003
A fuzzy logic expert system has been developed that automatically allocates electronic attack resources on different platforms in real-time. This resource manager is made up of four trees, the isolated platform tree, the multi-platform tree that allows an individual platform to respond to a threat. The multi-platform tree allows a group of platforms to respond to a threat in a collaborative fashion. A genetic algorithm is used to optimize the resource manager. A genetic program is used to evolve optimal fuzzy decision tree topology. The non-uniqueness of the tree structure is considered. The superiority of a genetic program evolved tree compared to a tree written down exclusively based on expertise is discussed. Fuzzy membership functions related to four fuzzy concepts are given. Experiments designed to test these concepts in the expert system are discussed, as well as the resource manager's ability to: allow multiple platforms to self-organize without the benefit of a commander; to tolerate errors made by other systems; and to deal with multiple distinct enemy strategies.