An Adaptive Decoupling Control Design for Flotation Column: A Comparative Study Against Model Predictive Control

Abstract

Multivariable nature of the flotation process compels advanced control strategies, including Model Predictive Controllers (MPC), leading to robust and effective compensation of disturbances. However, the complexity of calculations and implementation has limited their application in the industry. On the other hand, decoupling methods promise drastic reduction in computation time and complexity, while considering the multivariable relations between the inputs and outputs indirectly. Therefore, in this paper, an Adaptive Decoupling Controller (ADC) was designed and implemented on a pilot flotation column and its control performance was compared to MPC and SISO PID. Results showed that the weak coupling of variables in the studied case (and also reported by other researchers) causes a satisfactory control of the column by ADC, even better than the MPC in terms of Integral Squared Error index (ISE), Mean Squared Error index (MSE) and total energy consumption, with lower costs of computation and implementation.

KEYWORDS:

• Adaptive decoupling controller
• Adaptive systems
• Couplings
• Flotation column
• Predictive control
• Process modeling

Additional information

Notes on contributors

Sadreddin Nasseri

Sadreddin Nasseri was born in Tehran, Iran, in 1988. He received the BSc in mining engineering from Birjand University in 2010, and the MSc in mineral processing engineering from Tarbiat Modares University, Tehran, Iran, in 2014. He is a PhD scholar at Tarbiat Modares University, Tehran, Iran. His main research interests are observation and control in mineral processing. Email: [email protected]

Mohammad Reza Khalesi

Mohammad Reza Khalesi, born in Tehran in 1977, is a graduate of University of Tehran in mining exploration (BSc) and mineral processing (MSc) and holds a PhD in material and metallurgy from Laval University, Canada. He is an associate professor of Mineral Processing in Tarbiat Modares University, Iran; and a member of Iranian Society of Mining Engineering and Canadian Institute of Mining, Metallurgy and Petroleum (CIM). He is mostly focused on modeling, optimization, and control of mineral processing and extractive metallurgy systems.

Amin Ramezani

Amin Ramezani (IEEE Member from 2005) was born in Khuzestan, Iran, in 1979. He received his BSc in electrical engineering from Shahid Beheshti University, Tehran, Iran, in 2001, and the MSc in control systems from Sharif University of Technology, Tehran, Iran, in 2004 and PhD degree in electrical engineering from the University of Tehran, Tehran, Iran, in 2011. In 2012, he joined the Department of Electrical Engineering, Tarbiat Modares University, Tehran, Iran, as a lecturer. His current research interests include fault tolerant control systems, model-based predictive control systems, stochastic control systems, and hybrid systems. Dr Ramezani has been a member of IEEE since 2005 and a member of Iranian Community of Instrument and Control Engineers since 2012. Email: [email protected]

Mahmoud Abdollahi

Mahmoud Abdollahy is a professor of Mineral Processing at Tarbiat Modares University (TMU) in Tehran, Iran. He holds a MSc degree in mining exploitation from Tehran University (1987) and a PhD degree from Leeds University (1996), England. He is teaching flotation, advanced flotation, ore processing, and Treatment of Leach Solutions for MSc students and chemistry of solutions for PhD students at mineral processing Department of TMU. Email: [email protected]

Mehdi Mohseni

Mehdi Mohseni is an assistant professor of Mineral Processing at Tarbiat Modares University. He holds a BSc degree in mining exploitation from Sahand University of Technology (2008) and MSc and PhD in mineral processing from Tarbiat Modares University (2011 and 2017, respectively). Flotation and processing of the metallic and industrial minerals are his research and practical interest, along with molecular modeling and automation of mineral processing systems. Email: [email protected]