Improving Performance of Disturbance Rejection for Nonlinear Systems Using Improved Equivalent-Input-Disturbance Approach

Abstract

Disturbance rejection for industrial control systems with nonlinearities is a challenge. A disturbance-rejection control paradigm has been proposed to handle disturbances and nonlinearities by simply treating them as a total disturbance. However, this simplification has performance limitations as the characteristics of disturbances and nonlinearities are usually different. To this end, an extra degree of freedom is devised in this study to separate the estimation of an unknown disturbance and a system nonlinearity based on the equivalent-input-disturbance (EID) approach. Two scenarios, i.e., the nonlinearity with and without a known mathematical model, are considered. The information on the nonlinearity (for the scenario with the known model) or an extended state (for the scenario without the known model) is employed to design an observer and an EID estimator. The estimate of the nonlinearity is calculated using the mathematical model or the extended state, while the estimate of the disturbance is obtained from the EID estimator. An inner loop is then constructed to counteract the effect of them by using the estimates. A composite control law is devised to improve control performance by using feedforward control, feedback control, and inner-loop compensation. Stability criteria are derived using the small-gain theorem. The bounds of the errors in estimating the nonlinearity and the disturbance are analyzed. Simulation and experimental results demonstrate the validity of the method.