Control of Fully Actuated Systems With Perturbed Input Matrices and Partial Knowledge of Uncertainty Bounds

A robust adaptive control problem of fully actuated systems (FASs) with perturbed input matrices, unknown time-varying parameters, and partial knowledge of uncertainty bounds is considered. Unlike the existing results on FASs with perturbed input matrices, the restriction that the bound of the nonlinear uncertainty must be known is removed and a certain unknown constant is now allowed to be included in the bound. A novel robust adaptive controller is developed, where the adaptive term adapts to the effects of the time-varying unknown parameters in the FAS and the unknown constant in the uncertainty bound, and the robust term overcomes the influence of the remaining uncertain part caused by the perturbed input matrix. The designed controller enables the states of the closed-loop system and the estimation errors to be globally bounded. Finally, the developed method is successfully applied to the control of both a numerical example and an electromechanical system, and the effectiveness is proven by the comparative simulation results with the existing methods.