Data-Driven Iterative Learning Control for I/O Constrained LTI Systems

Abstract:In this paper, a novel data-driven iterative learning control(ILC) scheme is proposed for a class of constrained linear time-invariant(LTI) discrete-time systems with unknown system matrix. First, we only use the input and output information from previous iterations to estimate system Markov matrix obtained by the lifted technique. Second, three types of constraints,which include input constraint, system output constraint and the constraint of change rate of control input between two iteration index, are transferred into a linear constraint inequality in a unified manner. And then, by using the constructed objective function together with the linear constraint inequality, a quadratic programming(QP) problem can be formulated. The ILC control law in numerical manner can be obtained by solving the QP problem. The proposed control scheme is data-driven and does not need any model information. An illustrative example is provided to demonstrate the effectiveness of the proposed data-driven ILC strategy.

Translation or Not:no