A Neural Network Parallel Adaptive Control Algorithm for Fighter Aircraft Pitch Rate Control

摘要

This paper proposes a fighter aircraft pitch rate command-tracking algorithm based on a Neural Network Parallel Adaptive Controller (NNPAC). The control scheme consists of an online growing Radial Basis Function Neural Network (RBFNN) in parallel with a direct Model Reference Adaptive Control (MRAC) law. The RBFNN structure including centers, widths and radii are updated online based on a novel concept of an Intelligent Supervisory Loop (ISL) approach for accurate pitch rate command tracking. This update is carried out in view of the error between the pitch rate command and the output, in addition to a direct MRAC law which tracks the desired reference model output. The total control output is used to change the elevator deflection keeping the other control surface deflections at random values, even when the aircraft operates at difference maneuvers. Moreover, a suitable reference model structure is used for all aircraft operating modes and the system is then fully tuned by the parallel controller. The importance of the proposed scheme is in its ability to perform effectively as the system changes its modes. The proposed control algorithm has been validated by conducting simulation studies on a nonlinear F16 fighter aircraft model operating at different modes with randomly changing parameter set.