Talking MPC with Dr Brooks

Talking MPC with Dr Brooks

The Department of Mathematical Sciences in the Faculty of Science, Agriculture and Engineering recently hosted a webinar during which Dr Kevin Brooks gave a talk entitled: “All models are wrong but some are useful – Model Predictive Control (MPC) in Industry”.

From the Department of Mathematical Sciences, Dr Syamala Krishnannair took the initiative to arrange the webinar, in her capacity as the EXCO of the South African Council for Automation and Control (SACAC) and as an active researcher in automation and control. She invited Dr Brooks to deliver the talk. She also chaired the online event. According to her, the purpose of the webinar was to expose students in the faculty, and particularly students in the Department of Mathematical Sciences, to ‘automation and control’ as a potential field of study and employment opportunities offered by the industries. Dr Krishnannair also delivered a presentation describing the activities of SACAC. She attested that the students, specifically those in her research group, immensely benefited from the talk.

The Acting Dean of the Faculty of Science, Agriculture and Engineering, Prof Khoboso Lehloenya, in her welcome address, thanked Dr Brookes for availing himself to be the keynote speaker. “This talk came at a time when it is needed during this pandemic, Covid-19, when the world economy is under stress and struggling,” she acknowledged.

Dr Brooks’ talk was arranged under the auspices of SACAC. While he is employed at HATCH, from July 2019 he was appointed visiting adjunct professor in the school of chemical and metallurgical engineering at Wits. Here, he assists with the teaching of undergraduate process control and supervision of students in the DANCE (Data Analytics, Numeric and Control Engineering) group.

This webinar was an excellent opportunity for Dr Brooks to present and also enlighten on what the Model Predictive Control (MPC) is and how it is used to control a process while satisfying a set of constraints. It has been in use in the process industries in chemical plants and oil refineries since the 1980s. In recent years, it has also been used in power system balancing models and in power electronics. Model predictive controllers rely on dynamic models of the process and most often rely on linear empirical models obtained by system identification.

The principles and benefit of MPC was illustrated in his talk using the mechanism of ‘shower’ as an example. The theoretical background of MPC using a linear dynamic model and its application in prediction, identification and control were discussed in detail. He also solved a control problem (a case study from grinding mill) using optimisation.

Amongst the attendees was the Head of the Department of Mathematical Sciences Dr Maba Matadi, who highlighted the advantages of MPC. “The main advantage of MPC is the fact that it allows the current timeslot to be optimised, while keeping future timeslots in account. This is achieved by optimising a finite time-horizon, but only implementing the current timeslot and then optimising again, repeatedly, thus differing from Linear-Quadratic Regulator (LQR). Also, MPC has the ability to anticipate future events and can take control actions accordingly.

“A proportional–integral–derivative (PID) controllers do not have this predictive ability. MPC is nearly universally implemented as a digital control, although there is research into achieving faster response times with specially designed analogue circuitry. In the industry we hardly do things because they are interesting. There has to be a monetary return for better control. The monetary return for better control is for better control because of its less variation. In your system you can operate close to the constraint of the system,” explained Dr Matadi

In his closing remarks, Dr Matadi thanked all attendees and all those who worked tirelessly behind the scenes, more especially the guest speaker Dr Brookes for his informative talk.

  • Precious Shamase

Leave a Reply

Your email address will not be published. Required fields are marked *