Grant: Alliance Grants
McMaster University
$40,000.00
Reinforcement Learning Control of Distillation Towers: Avoiding Tower Flooding
FY 2023-24
Summary
Recent decades have seen a tremendous growth in sensing and data transfer capabilities, along with a parallel growth in machine learning/AI capabilities. The majority of the applications of AI have been in image recognition and classification (or even in generative language processing tools such as ChatGPT). As of yet, these techniques have not been utilized directly to run, or assist process operation, and to achieve desirable process behaviour, for instance to achieve flooding free control of distillation towers. This topic is of specific interest to the industrial partner of the project, but also to oil refineries in general. The PI is uniquely positioned in that he has expertise in data driven modeling and control and reinforcement learning (a machine learning technique) and a track record of working in close collaboration with industrial partners looking to turn the wide availability of data into meaningful information for the purposes of operating their processes better and more safely. This proposal will therefore develop the bring reinforcement learning techniques to process control in general, and distillation towers in particular, with potential applicability to several industrial sectors including energy systems, metal refining, bio-pharmaceutical production, and polymers processing. Since the method is inherently data driven, practitioners will be able to readily adapt them to improve their process operation in ways that will make the process operation more environmentally friendly and sustainable. The impact to Canada is manifold- first, it will widely impact Canada's goal of moving towards a sustainable economy and manufacturing and do it through a highly engaged industrial partner. Second, the involvement of the PI with the McMaster Advanced Control Consortium will mean that the methods will benefit widely the industrial sectors that are either already core to (such as metal refining), or expected to be, a central part of Canada's manufacturing sector (such as biopharmaceuticals). Last, but not the least, it will create HQP in areas needed by Canada.
Program
Alliance Grants
Recipient
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
591004-2023
Name nom
Mhaskar, Prashant P
Department département
Chemical Engineering
Organization id
27
Institution établissement
McMaster University
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2023
Competition year année de concours
2023
Award amount
40000
Program id
ALLRP
Program name en
Alliance Grants
Program name fr
Subventions Alliance
Group en
RESEARCH PARTNERSHIPS
Group fr
PARTENARIATS DE RECHERCHE
Committee code
96
Committee name en
RPP Internal Decision Cttee
Committee name fr
Cté de décision interne - PPR
Area of application code
500
Area of application group en
Manufacturing processes and products
Area of application group fr
Procédés et produits industriels
Area of application en
Manufacturing processes and products
Area of application fr
Procédés et produits industriels
Research subject code
9999
Research subject group en
Not available
Research subject group fr
Not available
Research subject en
Non disponible
Research subject fr
Non disponible
Application title
Reinforcement Learning Control of Distillation Towers: Avoiding Tower Flooding
Keyword
Data Driven Control; Model Predictive Control; Reinforcement Learning; Tower Flooding
Application summary
Recent decades have seen a tremendous growth in sensing and data transfer capabilities, along with a parallel growth in machine learning/AI capabilities. The majority of the applications of AI have been in image recognition and classification (or even in generative language processing tools such as ChatGPT). As of yet, these techniques have not been utilized directly to run, or assist process operation, and to achieve desirable process behaviour, for instance to achieve flooding free control of distillation towers. This topic is of specific interest to the industrial partner of the project, but also to oil refineries in general. The PI is uniquely positioned in that he has expertise in data driven modeling and control and reinforcement learning (a machine learning technique) and a track record of working in close collaboration with industrial partners looking to turn the wide availability of data into meaningful information for the purposes of operating their processes better and more safely. This proposal will therefore develop the bring reinforcement learning techniques to process control in general, and distillation towers in particular, with potential applicability to several industrial sectors including energy systems, metal refining, bio-pharmaceutical production, and polymers processing. Since the method is inherently data driven, practitioners will be able to readily adapt them to improve their process operation in ways that will make the process operation more environmentally friendly and sustainable. The impact to Canada is manifold- first, it will widely impact Canada's goal of moving towards a sustainable economy and manufacturing and do it through a highly engaged industrial partner. Second, the involvement of the PI with the McMaster Advanced Control Consortium will mean that the methods will benefit widely the industrial sectors that are either already core to (such as metal refining), or expected to be, a central part of Canada's manufacturing sector (such as biopharmaceuticals). Last, but not the least, it will create HQP in areas needed by Canada.