On active disturbance rejection control for nonlinear systems using time-varying gain.

In this paper, we propose an modified nonlinear extended state observer (ESO) with a time-varying gain in active disturbance rejection control (ADRC) to deal with a class of nonlinear systems which are essentially normal forms of general affine nonlinear systems. The total disturbance which includes unknown dynamics of the system, external disturbance, and unknown part of the control coefficient is estimated through ESO and is canceled in nonlinear feedback loop. The practical stability for the resulting closed-loop is obtained. It is shown that the “peaking value” occurred often in the constant high gain design can be significantly reduced by the time-varying gain approach. [ABSTRACT FROM AUTHOR]