INTERACTION MEASURES AND DECENTRALIZED CONTROL OF CHEMICAL-PLANTS

作     者：ZEGHAL, S PALAZOGLU, AN 

作者机构：ECOLE CENT PARIS LAB GIC F-92295 CHATENAY MALABRY FRANCE CTR RECH CHEM DIGUE BP 76 F-78600 MAISONS LAFFITTE FRANCE UNIV CALIF DAVIS DEPT CHEM ENGN DAVIS CA 95616 USA 

出 版 物：《COMPUTERS & CHEMICAL ENGINEERING》 (计算机与化工)

年 卷 期：1993年第17卷第Sup1期

页      面：S335-S341页

核心收录：

学科分类：0817[工学-化学工程与技术] 08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

主　　题：DYNAMIC SIMULATION PROCESS CONTROL INTERACTIONS PAIRING RELATIVE GAIN ARRAY BLOCK RELATIVE GAIN INTERNAL MODEL CONTROL MULTIPLE-INPUT MULTIPLE-OUTPUT SPEEDUP 

摘      要：Control system synthesis consists of two major subtasks: selection of measurements and manipulated variables, and deciding on the structure interconnecting the measured and manipulated variables along with the control law governing the interconnections. For multiple-input-mutiple-output systems, a decentralized control structure is generally preferred to a fully centralized structure. A major step in this controller design process is therefore the pairing problem: which manipulated variables to connect to which measurements? The answer to this question has undoubtedly to account for loop interactions that arise within the system. Work has been carried out using Relative Gain Array (RGA) and Block Relative Gain (BRG) interaction measures to develop automated pairing procedures and implement them as a portable software. A methodology involving the use of this software in conjuction with dynamic simulation and Internal Model Control design techniques for controllers was used on a particular case study to generate a satisfactory control structure.