Online detection of steady-state operation using a multiple-change-point model and exact Bayesian inference

ABSTRACT The detection of steady-state operation is critical in system/process performance assessment, optimization, fault detection, and process automation and control. In this article, we propose a new robust and computationally efficient online steady-state detection method using multiple change-point models and exact Bayesian inference. An average run length approximation is derived that can provide insight and guidance in the application of the proposed algorithm. An extensive numerical analysis shows that the proposed method is much more accurate and robust than currently available methods.