Development of a Nonlinear Reconfigurable F-16 Model and Flight Control Systems Using Multilayer Adaptive Neural Networks

For flight control systems, this paper proposes an adaptive control approach based on a framework of Explicit Model Following Direct Adaptive Control scheme. As a first step, a modified F-16 dynamics model is developed to explore control surface redundancies, as well as to enable modelling of dynamics changes result from faults, failures and/or plant deviations. In this modified model, each control surface can be individually controlled. Next, this paper proposes a flight control framework that integrates adaptive neural network, non-linear dynamic inversion, control allocation, and model reference following to leverage their synergies. Then, the proposed approach is tested using the F-16 nonlinear model developed and its performance is validated via numerical simulations.