Self Designing Flight Control System

Technical advances made in related topics now point the way for optimum continuous, on-line, real-time, self designing flight control systems. These advances include (1) algorithms using neural networks to enhance on-line parameter estimation and continuous control reconfiguration, (2) analytically derived optimum control for the coupled nonlinear inner loop and outer loop dynamics of high-performance flight vehicles, (3) synthesis of neural networks that initialize two-point boundary-value control solutions, (4) powerful airborne computing capabilities, and (5) new sensing techniques and adaptive materials. This proposal presents a unifying conceptual design framework for integration of these capabilities into self designing flight control systems to eliminate gain switching, accommodate failures and damage, maximize maneuverability and the flight envelope, and increase the probability of successful mission completion. In Phase I Barron Associates will refine this concept, assess its feasibility, practicality, and current technology limitations, develop critical analytical and computational tools, and perform preliminary simulations. In Phase II Barron Associates will implement the self designing controller and interface it with the simulation of a representative aircraft to demonstrate its operation in realistic scenarios.