Comparison of input-output pairing selection methods.

• Main Author: Chong, Chen Nee.
• Other Authors: Vinay Kumar Kariwala
• Format: Final Year Project
• Language: English
• Published: 2009
• Subjects: DRNTU::Engineering::Chemical engineering::Processes and operations
• Online Access: http://hdl.handle.net/10356/16523
• Institution: Nanyang Technological University

Decentralized control systems are attractive options for control of multivariable processes as compared to centralized control systems due to its design simplicity and failure tolerance. Depending on the plant size, a decentralized controller may consist of a few up to hundreds sets of single-input-single-output (SISO) sub-controllers. During the design of these SISO controllers, interactions among loops are assumed to be negligible. However, in the actual plant, loops interact with each other potentially leading to deterioration in closed-loop performance. Thus, it is important to pair inputs and outputs such that interactions are minimized. Different patterns of pairing/interconnections between the system’s inputs and outputs can lead to different control configurations. As plant size increases, the number of possible configurations grows exponentially. In order to save costs and computing time, it is desired to have systematic approaches to screen out unfavorable configurations for large-scale systems. Apart from interaction minimization, it is desired to choose configurations that meet certain criteria, such as integrity, robustness and nominal performance, to enable the design of a controller to achieve an ideal control system. There are many input-output (IO) pairing methods available in literature. The main objective of this thesis is to compare the ability of the different methods in satisfying the above-mentioned criteria. Qualitative assessments and comparisons of the reviewed methods are presented. In addition, a case study on a 3 x 3 system is conducted to obtain simulation results on the dynamic performance of each pairing alternative. Lastly, this thesis points out open issues and provides guidelines regarding the future research directions in IO pairing.