Observer-Based Sliding Mode Predictive Control for Piezoelectric-Driven Active-Passive Vibration Isolator With Variable-Mass Payloads

This study investigates the output feedback sliding mode predictive control (SMPC) method to address the microvibration suppression problem for piezoelectric-driven active-passive vibration isolator (PDAPVI) with variable-mass payloads. First, a neural network (NN)-based state observer is adopted to facilitate the output feedback controller design. Then, the generalized proportional integral observer (GPIO) is developed to estimate and predict the matched and mismatched disturbances. Based on the estimated information, a novel sliding mode surface is designed to address mismatched uncertainties and disturbances. Moreover, an integrator of sliding mode state error (ISMSE) is introduced to enhance the control performance for system with variable-mass payloads. By combining the predicted sliding mode state, predefined reference, and ISMSE, the optimal control signal is derived via the receding horizon optimization. Rigorous analysis is provided to prove the convergence of the closed-loop system and its capability to mitigate mismatched uncertainties and disturbances. Finally, experiments are conducted to validate the effectiveness of the proposed scheme.