Optimal Allocation of Distributed Resources Using Fuzzy Logic and a Genetic Algorithm

摘要

A fuzzy logic-based resource manager (RM) that will allocate electronic attack (EA) resources distributed across many platforms is under development. The platforms will consist of ships, helicopters, airplanes, land-based facilities, robotic devices, and, potentially, satellites. The RM will allow codification of military expertise in a simple mathematical formalism known as the fuzzy decision tree. The fuzzy decision tree will form what is known as a fuzzy linguistic description, i.e., a formal fuzzy if-then rule-based representation of the system. Since the decision tree is fuzzy, the uncertainty inherent in the root concepts propagates throughout the tree. The functional form of the fuzzy membership functions for the root concepts will be selected heuristically and will generally carry one or more free parameters. The free parameters in the root concepts will be determined by optimization both initially and later at noncritical times. A genetic algorithm will be used for optimization.