Parallel Nonlinear Adaptive Genetic Algorithm for Man-in-the-Loop Air-to-Air System Performance Evaluati

Physical Optics Corporation (POC) proposes to develop a parallel adaptive genetic algorithm module to evolve better tactics in real time for air-to-air engagements. The system will be based on a highly non-linear, fuzzified genetic algorithm that will be implemented in a distributed computer environment. It specifically has the following innovations: 1) a highly flexible, fast, and efficient decision making aid based on the integration of adaptive fuzzy logic and pre-filtering to control the genetic operators; 2) multiprocessor parallel computing; 3) a LAN-based system that allows simulations of many-verus-many (MvN) engagements. By implementing a LAN-oriented, multiplayer aerial combat simulation environment, POC will demonstrate that genetic algorithms can be used to solve real-life optimization problems. In Phase I, POC will investigate the feasibility of the proposed concept through algorithm layout and program coding, and will present a proof-of-concept demonstration at the end of Phase I.