Control structure selection for the ALSTOM gasifier benchmark process using GRDG analysis

Abstract

The objective of this work is to explore the disturbance rejection capability of possible multi-loop control structures for the ALSTOM gasifier benchmark process and selecting the appropriate control structure. Generalized Relative Disturbance Gain (GRDG) analysis is used for control structure determination. In order to carry out GRDG analysis, process models in the form of transfer functions are obtained from the discrete time models identified using the Output-Error (OE) method for system identification. Models identified with the OE method can provide accurate long range prediction (or simulation) performance and, hence, lead to accurate transfer function models. The GRDG analysis results clearly show that the baseline controller proposed by Asmar et al. (Asmar et al., 2000) is the favoured multi-loop control structure among their initial designs. This study provides explanation for the reason behind the impressive performance of the ALSTOM baseline controller that is not available before.

Performance evaluation of optimal PI controller for ALSTOM gasifier during coal quality variations

“…Researchers have attempted to design controllers and/or retuned the baseline controller to meet the performance objectives at 100%, 50% and 0% load conditions. While many researchers have attempted these challenges [3][4][5][6][7][8][9][10][11][12][13][14][15][16], meeting the desired gasifier performance due to variation in calorific value under step and sinusoidal pressure disturbance remains unsolved. The ultimate requirement is to design an optimal controller such that all the constraints are met for disturbances around all load conditions and coal quality variations.…”

Optimal Tuning of Decentralized PI Controller of Nonlinear Multivariable Process Using Archival Based Multiobjective Particle Swarm Optimization

“…This benchmark challenge II, which is a higher order state space model, is reduced to lower order model by different techniques [12,13]. Design and implementation of advanced control schemes are also reported in the literature [14][15][16][17][18][19][20][21][22][23]. Soft computing techniques such as Bat algorithm, Cuckoo search, Nondominated Sorting Genetic algorithm II, Multiobjective Genetic algorithm, and Normalized Normal Constraint algorithm are also found in the literature [6,7], [22][23][24][25][26][27][28], which deals with tuning of baseline PI controller.…”

Branch and bound method for multiobjective pairing selection

“…In some cases, pairings can be selected based on process knowledge and heuristics. For interacting multivariable processes often encountered in industries [1], however, systematic tools are needed to complement the engineering insights. During the past few decades, a number of useful tools have been developed for selection of appropriate pairings [31,33].…”

“…For efficiently selecting pairings based on a single criterion, genetic algorithms [13] and mixed integer linear programs (MILP) [18] have been used. Different criteria, however, address different properties of the closed- 1 A preliminary version of this work was presented at 17th IFAC world congress held in Seoul, Korea [19]. 2 Corresponding Author: Tel: +65-6316-8746; Fax: +65-6794-7553 ; E-mail: vinay@ntu.edu.sg loop system and are often conflicting in nature [5].…”