INFERENTIAL CONTROL BASED ON A MODIFIED QPLS FOR AN INDUSTRIAL FCCU FRACTIONATOR

Abstract

A modified Quadratic Partial Least Squares (MQPLS) algorithm based on nonlinear constrained programming is proposed. Sequential Unconstrained Minimization Technique (SUMT) is employed to calculate the outer input weights and the parameters of inner relationship. It was found that MQPLS can not only explain more of the underlying variability of the data, but also has improved modelling and predictive ability. An inferential control system is implemented on the Distribute Control System (DCS) of a fluid catalytic cracking unit (***FCCU) main fractionator. A soft sensor MQPLS–based was developed to estimate solidifying point of diesel oil. The controller was established via constrained Dynamic Matrix Control (DMC) algorithm. Real time application results demonstrated the performance of the inferential control system based on MQPLS was much better than the original tray temperature control system. This resulted in a 1.0% increase in production rate, and a significant increase in profit.