A Tracking Robust Learning Control for Micro Scale Actuator Systems

Abstract

In this paper, a robust sliding mode-based learning control (RSMLC) scheme is developed for a micro-scale actuator (MSA) system. By exploiting the instantaneous information of the system closed-loop stability, a sliding mode controller based on the recursive learning technique is designed. The proposed control approach can govern the sliding variable and tracking error between the desired references and actual displacement of the MSA to asymptotically converge to zero. Unlike conventional sliding mode control (CSMC), no prior knowledge of MSA system uncertainties and parameters variation is required in the RSMLC design. Besides, there is no explicit switching element in the RSMLC structure, it can be considered as a chattering free control method. Meanwhile, the inherent high robustness property of the CSMC is fully preserved in the proposed controller. Where the learning algorithm continuously adjusts the closed-loop response based on the most recent history of closed-loop stability. Thus, the stability and convergence analysis of the RSMLC is proved rigorously. Simulation studies have been conducted for a piezoelectric actuator (PEA) system, as a prototype of MSA. For comparison reasons, these results are presented correspondingly with CSMC results to demonstrate the effectiveness of the proposed controller over the CSMC.