Grant: Discovery Launch Supplement
University of Alberta
$12,500.00
Machine learning based control in transportation and smart agriculture: enhancing performance, efficiency and robustness
FY 2024-25
Summary
MOTIVATION The Canadian Net-Zero Emissions Accountability Act targets net-zero greenhouse gas (GhG) emissions by 2050; however, as the second largest country by area, the transportation and agricultural sectors contributed 32% of the total GhG emissions in 2021 leading to the need to maximize efficiency and reduce emissions. The reliability and high energy density of internal combustion engines (ICEs) has yet to be matched by battery-electric vehicles. A hydrogen fuel cell (FC) can provide zero-emission electricity for an electric motor. Alternatively, in the short-term, the GhG emissions from ICEs can either be reduced or eliminated with zero-carbon fuels such as hydrogen or ammonia. In both cases, optimal control strategies are needed for the best performance and lifespan from the energy conversion systems, which convert chemical energy into useful mechanical power. Model predictive control (MPC) is one of the most promising control strategies for handling these highly constrained nonlinear systems. The proposed program will focus on integrating machine learning (ML) for the model and controller to discover state of the art control methods to optimize energy conversion in mobile applications.
PROGRAM OBJECTIVES AND APPROACHES The long-term objective of the program is to develop and experimentally validate ML-based control strategies to optimize the performance and efficiency of energy conversion systems used in transportation and agriculture systems. This program looks to understand and develop ML approaches that when combined with traditional control strategies provide effective control of various power conversion systems and understand their limitations. Building on the recent advances in ML, MPC and the computational performance of embedded hardware, the short-term objectives of this program are: (1) experimental and physics-based modeling of ICE's using zero-carbon fuels and proton-exchange membrane FC (PEMFC), (2) ML-based modelling of the ICEs to predict engine performance and emissions during transient operation and modelling of performance, hydration and health of PEMFC, and (3) improving experimental MPC robustness through use of state estimation and MPC behavior cloning using ML for embedded control hardware.
IMPACT OF PROPOSED RESEARCH Environment: The proposed research outcomes will have a direct impact on the environment by providing a solution to eliminate the emissions by replacement with a FC and to improve the efficiency of ICEs when operated with zero-carbon fuels. Economic Growth: Reducing the complexity of ML-based MPC development will make ML accessible to small/mid-sized companies. Additionally, increasing the use of zero-carbon fuels in ICE and fuel cell systems will allow the production of these fuels to grow allowing Canada to become a leader in producing and handling these fuels. Social Equity: Reducing emissions and improving air quality will have a direct benefit on the quality of life and health of Canadians.
Program
Discovery Launch Supplement
Recipient
Location
Alberta, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
DGECR-2024-00341
Name nom
Gordon, David
Department département
Mechanical Engineering
Organization id
9
Institution établissement
University of Alberta
Province en
Alberta
Province fr
Alberta
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
12500
Program id
DGECR
Program name en
Discovery Launch Supplement
Program name fr
Supplément Tremplin vers la découverte
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1512
Committee name en
Mechanical Engineering
Committee name fr
Génie mécanique
Area of application code
516
Area of application group en
Manufacturing processes and products
Area of application group fr
Procédés et produits industriels
Area of application en
Mechanical machinery, heavy equipment (including farm, forestry, and construction equipment)
Area of application fr
Machinerie mécanique, machinerie lourde (y compris machinerie agricole, forestière, de construction)
Research subject code
2100
Research subject group en
Mechanical engineering
Research subject group fr
Génie mécanique
Research subject en
Mechanical engineering
Research subject fr
Génie mécanique
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Machine learning based control in transportation and smart agriculture: enhancing performance, efficiency and robustness
Keywords
machine learning; model predictive control; energy conversion systems; combustion engines; realtime control; process modeling; sensors; process modeling; fuel cells; mechatronics
Application summary
MOTIVATION The Canadian Net-Zero Emissions Accountability Act targets net-zero greenhouse gas (GhG) emissions by 2050; however, as the second largest country by area, the transportation and agricultural sectors contributed 32% of the total GhG emissions in 2021 leading to the need to maximize efficiency and reduce emissions. The reliability and high energy density of internal combustion engines (ICEs) has yet to be matched by battery-electric vehicles. A hydrogen fuel cell (FC) can provide zero-emission electricity for an electric motor. Alternatively, in the short-term, the GhG emissions from ICEs can either be reduced or eliminated with zero-carbon fuels such as hydrogen or ammonia. In both cases, optimal control strategies are needed for the best performance and lifespan from the energy conversion systems, which convert chemical energy into useful mechanical power. Model predictive control (MPC) is one of the most promising control strategies for handling these highly constrained nonlinear systems. The proposed program will focus on integrating machine learning (ML) for the model and controller to discover state of the art control methods to optimize energy conversion in mobile applications.
PROGRAM OBJECTIVES AND APPROACHES The long-term objective of the program is to develop and experimentally validate ML-based control strategies to optimize the performance and efficiency of energy conversion systems used in transportation and agriculture systems. This program looks to understand and develop ML approaches that when combined with traditional control strategies provide effective control of various power conversion systems and understand their limitations. Building on the recent advances in ML, MPC and the computational performance of embedded hardware, the short-term objectives of this program are: (1) experimental and physics-based modeling of ICE's using zero-carbon fuels and proton-exchange membrane FC (PEMFC), (2) ML-based modelling of the ICEs to predict engine performance and emissions during transient operation and modelling of performance, hydration and health of PEMFC, and (3) improving experimental MPC robustness through use of state estimation and MPC behavior cloning using ML for embedded control hardware.
IMPACT OF PROPOSED RESEARCH Environment: The proposed research outcomes will have a direct impact on the environment by providing a solution to eliminate the emissions by replacement with a FC and to improve the efficiency of ICEs when operated with zero-carbon fuels. Economic Growth: Reducing the complexity of ML-based MPC development will make ML accessible to small/mid-sized companies. Additionally, increasing the use of zero-carbon fuels in ICE and fuel cell systems will allow the production of these fuels to grow allowing Canada to become a leader in producing and handling these fuels. Social Equity: Reducing emissions and improving air quality will have a direct benefit on the quality of life and health of Canadians.