Metrics for Interaction Assessment in Multivariable Control Systems Using Directional Analysis

• Published in: Industrial & engineering chemistry research 2018-01, Vol.57 (3), p.967-979
• Main Authors: Garg, Abhinav, Tangirala, Arun K
• Format: Article
• Language: English
• ISSN: 0888-5885; 1520-5045
• DOI: 10.1021/acs.iecr.7b03671

This work presents a novel method for assessing and quantifying the level of interactions in multivariable control systems in a hierarchical manner. The proposed metrics are based on the total directed (causal) power transfer between a pair of variables, constructed from a jointly linear stationary representation of the process. Three prime features of these metrics, specifically, (i) the ease of interpretation and versatility in accommodating different controller structures and operating conditions, (ii) the ability to use both first-principles linearized and data-driven (empirical) models, and (iii) the means for quantifying and evaluating the suitability of a decentralized versus a centralized control scheme, make them highly practical and valuable in the design and assessment of multivariable control schemes. An important outcome of this work is also an operator-friendly visual tool for inspection of interactions in various loops that can be generated for different tuning methods and controller configurations. Simulation studies of four different benchmark processes are presented to demonstrate the efficacy of the proposed method.