An ARX-based Technique for Steady-State Identification of Chemical Processes

Steady-state identification is employed as part of different methods in process systems engineering, such as real-time optimization and advanced process control. In this paper, we analyze different steady-state identification techniques for chemical processes. In particular, a new method based on the ARX (Auto-Regressive model with eXogenous inputs) technique is proposed and compared with other methods available in the literature via two examples. The main advantages of the novel method lie on the design of its index, which considers the singularity of the model matrices, the reduction of the number of parameters to be tuned, and the intuitive selection of its threshold.