A fuzzy identification algorithm for NARMAX model and its application to the estimation of oxygen concentration

The continuous catalytic reforming system is an important process in petroleum refineries, and the estimation of oxygen concentration plays an important role in refinery enterprises. In this paper, based on the Takagi-Sugeno fuzzy inference system, a NARMAX model is built for a multi-input single-output system of oxygen concentration in a catalytic reformer. 21 variables are chosen as secondary variables related to oxygen concentration for the continuous catalytic reforming system, and PCA is used to treat the sample data to build the fuzzy model in order to reduce some redundant information. The practical operational data is used to build the estimation model of oxygen concentration, and the simulation results are very satisfactory.