Computers and Chemical Engineering 36 (2012) 265–272 Contents lists available at SciVerse ScienceDirect Computers and Chemical Engineering jo u rn al hom epa ge : www.elsevier.com/locate/compchemeng Comparative control study of a simulated batch rectifier Amiya K. Jana ∗ , Raja Sinha Department of Chemical Engineering, Indian Institute of Technology-Kharagpur, Kharagpur 721 302, West Bengal, India a r t i c l e i n f o Article history: Received 24 June 2010 Received in revised form 8 September 2011 Accepted 14 September 2011 Available online 22 September 2011 Keywords: Batch rectifier Simulation Luenberger-like nonlinear estimator Feedback linearizing controller a b s t r a c t This article proposes a nonlinear control strategy for a batch rectifier. The hybrid control strategy com- prises a state estimator, namely Luenberger-like nonlinear estimator (LNE) and the feedback linearizing controller (FLC). The observation scheme aims to estimate the imprecisely known parameter based on the available measurement. For the representative distillation unit, the state predictor is formulated using only a component material balance equation around the condenser–reflux drum system. It clearly indi- cates the existence of a process/model mismatch and this discrepancy is efficiently taken care of by the corrector part of the LNE. In the subsequent phase, the FLC is synthesized for the example system. The derived nonlinear FLC–LNE controller shows better performance compared to a traditional proportional integral (PI) law. The simple structure, straightforward design and good performance make this nonlinear controller attractive for online use. © 2011 Elsevier Ltd. All rights reserved. 1. Introduction Batch processing is becoming more important due to the rapid expansion of fine chemical, pharmaceutical and food industries. Batch operations are widely applied in these industries because of the low amount and the frequent change of the products. With- out using a train of continuous columns, one single batch rectifier has the ability to separate a multicomponent mixture into several products at a much lower capital cost. Now-a-days, the product purity requirements are usually strict and a significant economic penalty exists when the specification requirements are not met. Therefore, it is necessary to develop a correct methodology for tight composition control. Frattini-Fileti and Rocha-Pereira (1997) formulated the predictive and adaptive control laws to deal with the time-varying nature of the batch dis- tillation. Barolo and Berto (1998) derived a state estimator based nonlinear internal model control. To compute the top product composition from selected tray temperature measurements, they designed the extended Luenberger observer. A nonlinear model predictive controller was synthesized by Dechechi, Luz, Assis, Maciel, and Maciel Filho (1998) for distillate composition regula- tion. Alvarez-Ramirez, Monroy-Loperena, Cervantes, and Morales (2000) investigated the performance of a classical PID controller with antireset windup and an observer. In the subsequent period, Oisiovici and Cruz (2001) proposed an inferential control structure for the multicomponent batch columns using an extended Kalman filter. Han and Park (2001) used the quasi-dynamic estimator to ∗ Corresponding author. Tel.: +91 03222 283918; fax: +91 03222 282250. E-mail address: akjana@che.iitkgp.ernet.in (A.K. Jana). compute the distillate composition in a closed-loop batch recti- fier. Then, Li and Wozny (2001) proposed an adaptive controller, in which, a recursive least-square estimation with a variable forget- ting factor is used. Monroy-Loperena and Alvarez-Ramirez (2003) presented a method for the identification and nonlinear control of a batch distillation. Recently Gundale and Jana (2007) reported an adaptive estimator-based generic model controller for a multicom- ponent batch rectifier. Batch distillation is a nonlinear, complex and high-order separa- tion process. It is intrinsically dynamic, making the control a more challenging task. Composition profiles and operating conditions may change over a wide range of values during the entire batch operation and the state estimators can effectively deal with the nonstationary nature of the batch rectifiers (Mujtaba & Macchietto, 1996). Most of the work reported the nonlinear state estimation and control of continuous distillation systems. Relatively few arti- cles address the issue of estimator-based nonlinear control of batch rectifiers. It is with this intention that the present work has been undertaken. The main contributions of this work are highlighted below. (i) It is realized that most of the nonlinear state estimation algorithms involve complex and lengthy formulation. Originally, the large pre- dictor model and complex structure of the closed-loop observer complicate the overall design. In this contribution, the Luenberger- like nonlinear estimator (LNE) (Biagiola & Figueroa, 2004) has been developed for the example system. Keeping the issue of observer design complexity in mind, the LNE is formulated such a way that the predictor model includes only component mole balance equa- tion around the condenser–reflux drum envelope and an extra state equation having no dynamics. It is interesting to note that the pro- posed estimation approach only computes the states as per the 0098-1354/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.compchemeng.2011.09.013