Autocovariance-based plant-model mismatch estimation for linear model predictive control

• Published in: Systems & control letters 2017-06, Vol.104, p.5-14
• Main Authors: Wang, Siyun, Simkoff, Jodie M., Baldea, Michael, Chiang, Leo H., Castillo, Ivan, Bindlish, Rahul, Stanley, David B.
• Format: Article
• Language: English
• Subjects: Autocovariance; Model predictive control; Plant-model mismatch
• ISSN: 0167-6911; 1872-7956
• DOI: 10.1016/j.sysconle.2017.03.002

In this paper, we present autocovariance-based estimation as a novel methodology for determining plant-model mismatch for multiple-input, multiple-output systems operating under model predictive control. Considering discrete-time, linear time invariant systems under reasonable assumptions, we derive explicit expressions of the autocovariances of the system inputs and outputs as functions of the plant-model mismatch. We then formulate the mismatch estimation problem as a global optimization aimed at minimizing the discrepancy between the theoretical autocovariance estimates and the corresponding values computed from historical closed-loop operating data. Practical considerations related to implementing these ideas are discussed, and the results are illustrated with a chemical process case study. •An autocovariance-based plant-model mismatch estimation approach is proposed.•Explicit relations between closed-loop data statistics and mismatch are established.•Changing of constraint active sets in the MPCs are considered in the approach.•Estimates are very close to their true values in the case study.