Cascade predictor for a class of mechanical systems under large uncertain measurement delays

The state reconstruction for a class of mechanical systems with large uncertain measurement delays is investigated in this paper. A new cascade predictor which is composed of several sub-systems in a chain is proposed. The delayed system states are estimated in the first place by a generalized super-twisting sub-observer. Then the following sub-predictors are to predict the estimation results of the previous one under novel time-varying prediction horizons. The estimations of the actual states of the systems are obtained from the last sub-predictor. The boundness of prediction errors is proved through the Lyapunov functional method, and the convergence regions are determined by the uncertain parts of the system. Finally, several comparative simulation and experiment results are presented to verify the effectiveness of the proposed predictor. •Cascade predictor for mechanical systems with large measurement delays.•Novel prediction horizons are designed to deal with time-varying delays.•Robustness to uncertain delay is improved by using super-twisting observer.