Robust control of linear systems with nonlinear uncertainties via disturbance observer techniques

This paper considers robust control problems for linear systems with matched nonlinear uncertainties. In this paper, the uncertainties are considered as disturbances and a disturbance observer is designed. By subtracting these estimated values of disturbances from the plant input, the performance degradation due to the uncertainties can be improved. The estimation error is shown to be bounded and made arbitrarily small if the estimator gain is chosen large enough. To get rid of the influences of these errors on the performance, the additional robust min-max control input is added with the disturbance observer output. By the way, the min-max control input needs to be switched and provokes chattering phenomena. The control method suggested in this paper can, however, reduce chattering phenomena as small as possible by choosing large disturbance observer gain because the magnitude of the switching control input is determined by the ones of the disturbance estimation error bounds.