Robust ILC design for LPV systems incorporating preview repetitive control

Abstract

This paper proposes a novel mixed preview repetitive iterative learning control scheme for discrete-time linear parameter-varying (LPV) systems in linear fractional representation (LFR), aiming at high-precision tracking of periodic reference signals. The key innovation is the integration of preview control, repetitive control, and iterative learning within a unified LPV-LFR framework, enabling simultaneous use of future reference information and adaptation to parameter variations. An augmented error system is constructed to incorporate previewed reference data, and robust stability and convergence conditions are derived via linear matrix inequalities (LMIs) using parameter-dependent Lyapunov functions and slack variables. The effectiveness of the proposed method is demonstrated through simulations on two LPV systems, showing significant tracking error reduction over iterations. The PRI-ILC framework provides a systematic and robust solution by combining feedforward, repetitive, and learning-based control mechanisms.