Intelligent dynamic surface control strategy for high-performance aircraft subject to uncertain aerodynamics

The uncertain dynamics and actuator saturation in attitude tracking of high-performance aircrafts have posed a serious challenge in deriving an accurate control strategy. This paper presents an adaptive observer-controller strategy to deal with these problems. It uses wavelet neural networks (WNN) to estimate the functional uncertainties in the nonlinear dynamics of the aircraft with 6 degree of freedom (6 DoF). WNN uses wavelets as activation function to achieve superior learning characteristics. This system is very close to the real time model as it is subjected to the modelling uncertainties and actuator saturation owing to the external random parameters which makes the control design very complicated. The novelty of this paper lies within the estimation of these highly random nonlinear uncertainties by the modified deep learning network, WNN and the respective controller-observer strategy. Simulation analysis has been performed to evaluate the performance of the theoretical development presented in the paper.