A Data-Driven Closed-loop Subspace-aided Approach to Predictive Controller Design

A novel data-driven closed-loop subspace-aided approach to predictive controller design is proposed in this paper. The basic idea of the proposed method is to construct an instrumental variable which contributes to extending the standard LQ factorization to the closed-loop situation. Moreover, the predictive controller can be directly obtained from the input/output data with the subspace-aided technique by minimizing the normal predictive objective function. The efficiency of the proposed method is eventually verified and demonstrated through a numerical study on a randomly generated three-order discrete-time linear time invariant (LTI) system.