Parameter Selection and Performance Analysis of Linear Disturbance Observer Based Control for a Class of Nonlinear Uncertain Systems

The article presents a principle of selecting the parameters in linear disturbance observer based control (LDOBC). Via designing the numerator of Q-filter based on the presented principle, a class of optimal parameterized LDOBCs are proposed, which is featured with stronger robustness to uncertainties. Based on a new realization of proposed LDOBC and Lyapunov method, this article rigorously investigates the closed-loop transient performance under a wide class of nonlinear uncertainties described in time domain. Moreover, the relationship between the control parameters and the bounds of tracking and estimating errors is explicitly shown, which illuminates the highly consistent closed-loop performance of the presented LDOBC.