Predictive ESO-Based Anti-Disturbance Control Design for a Class of Nonlinear Systems With Time-Varying Input Delay

In this article, a predictive extended state observer (PESO)-based anti-disturbance control design is proposed for a class of nonlinear systems with time-varying input delay subject to typically encountered disturbances like constant, asymptotically stable, step, ramp or harmonic type in practice. In case that the external disturbance could be exactly described by an exogenous model, a novel nonlinear sequential PESO (NSPESO) is first introduced to asymptotically estimate the dynamics of the future system state and external disturbance. Based on the predicted state and disturbance, a nonlinear anti-disturbance controller is then designed to eliminate the adverse effect of external disturbance. If the external disturbance could not be completely described, another NSPESO-based anti-disturbance control scheme is further developed by augmenting the unmodeled disturbance dynamics as an extended state of NSPESO for estimation and counteraction, so as to improve the disturbance attenuation performance. Exponential stability of the resulting closed-loop systems is rigorously analyzed with proofs. It is revealed that the NSPESO gains are constrained by the upper bound of time delay under a fixed number of NSPESO. Finally, two illustrative examples are given to demonstrate the theoretical analysis and advantages of the proposed two control schemes over the existing methods.