Dynamic and Adaptive Control of Monoaxial GS with GAR. Part 2: Control Architecture Design and Validation

This paper presents the design of adaptive control subsystems of the precession angle of the gyroscope-inner gimbal assembly and of the angular velocity of rotation of the platform (outer gimbal) from the monoaxial force GS component with GAR. The nonlinear dynamics of the two stabilization and guidance subsystems were obtained in the first part of this paper series. The design of the two subsystems is based on the use of the concept of dynamic inversion and neural networks. The control structures include reference models, dynamic compensators (P.I. and, respectively, P.D.), linear state observers and neural networks. Neural networks model the adaptive components of the control laws, these having the role of compensating dynamic inversion errors. By numerical simulation, using Matlab/Simulink models of the two subsystems, the dynamic GS characteristics for stabilization and orientation regimes are drawn.