A Novel Algorithm for Model-Plant Mismatch Detection for Model Predictive Controllers

摘要

For Model Predictive Controlled (MPC) applications, the quality of the plant model determines the quality of performance of the controller. Model Plant Mismatch (MPM), the discrepancies between the plant model and actual plant transfer matrix, can both improve or degrade performance, depending on the context in which performance is measured. In this paper, we do not use performance metrics or “yes-no”-type tests to merely diagnose the presence or absence of MPM in the plant matrix. Rather, we achieve the further goal of locating the exact MPM-affected elements within the plant matrix. Our proposed detection algorithm consists of two system identification experiments: the first experiment diagnoses the presence of MPM, and the second experiment pinpoints the exact MPM-affected elements. We then exercise the algorithm on artificial 3x3 and 5x5 plants suffering from sparse MPM, and demonstrate the algorithm's capability of correctly locating the MPM-affected entries.