Prediction-based Adaptive Robust Control for a Class of Uncertain Time-delay Systems

Abstract

This paper presents an integrated control design approach for a class of dynamical systems that satisfy a certain matching condition subject to known input time-delay, unknown parameters, and time-varying disturbances, simultaneously. A novel nonlinear predictor adaptive robust control (PARC) is proposed to track a desired state trajectory. The controller uses predictor-based model compensation to attenuate the effect of input time-delay, gradient type projection with prediction-based learning mechanisms to reduce the parameter uncertainties, and prediction-based nonlinear robust feedback to attenuate the effect of model approximation errors and disturbances, simultaneously. The controller guarantees a prescribed transient performance (with global exponential convergence) and final steady-state tracking error with an ultimate bound proportional to the time-delay, the disturbances, and the switching gain. The effectiveness of the proposed control design is illustrated with a simple tumor growth example.