摘要:AbstractThis paper presents the research on the closed-loop data-driven design of the process monitoring system. The core of the study is the closed-loop identification of the data-driven stable kernel representation (SKR) of the system. The proposed method takes into account the effects of feedback controller and the noise information in the closed-loop is directly extracted from a single LQ decomposition. The main results and the proposed methods are verified and demonstrated through randomly generated numerical examples.