Genetic Fuzzy System for Solving the Aircraft Conflict Resolution Problem

Abstract: This chapter discusses a genetic fuzzy logic approach to solving the aircraft conflict resolution problem. The complexity of the problem is increased by adding uncertainty to the velocity and maneuver parameters, which means the aircraft position at any instant will be within a region of uncertainty represented by a convex hull. The objective is to obtain conflict-free trajectories for the aircraft such that the total cost of maneuvers is minimized. A unique architecture is discussed, which consists of a hidden layer of neurons and a layer of fuzzy inference systems (FISs) that provide the final output. An artificial intelligence called EVE is used to train the system, and once it is trained, its capability is evaluated on a set of test scenarios. We compare the cost and the computational time of our approach with that obtained by directly applying the genetic algorithm (GA). The results show the effectiveness of our approach for obtaining quick nearoptimal solutions.