Predictive functional control (PFC) applied to an unstable system. An industrial application

摘要

This paper describes an industrial application of Predictive Functional Control (PFC), which belongs to a family of Model Predictive Control (MBPC) developed by Richalet and coworkers during the last decades. The case presented in this work consists on controlling the bottom level and the reboiler temperature of a stripper distillation column, which belongs to the naphtha hydrotreater (Hydrobon) plant located at REPSOL-YPF Plaza Huincul refinery. The Hydrobon plant objective is to eliminate the downstream naphtha reforming (Platforming) plant catalyst. Level and temperature control by zone strategy is implemented; it involves keep the level and temperature inside a specific allowable band with respect to the set point and the corresponding controller tuning parameters are adjusted depending on that. The project consisted in three main steps, process identification, off-line controller design and on-line commissioning. The controllers were implemented in the control processor of the Distributed Control System (DCS). This paper presents the details of the identification, and off-line design, as well as the results obtained after the control strategies commissioning.