Grant: Discovery Grants Program - Individual
University of Toronto
$46,000.00
Towards model reduction for efficient, reliable, and automated physics-based simulation of industry-scale problems
EF 2024-25
Résumé
With advances in computing hardware and algorithms, high-fidelity simulations based on partial differential equations (PDEs) have become an indispensable tool in the design, analysis, and operation of transportation and energy systems. Continued advances in simulation capabilities are crucial to design next-generation engineering systems that are more efficient, safer, more sustainable, and more economical.
My research program focuses on the development of efficient, reliable, and automated computational methods for PDEs. We wish to provide efficient and reliable PDE solutions to address challenging many-query and real-time tasks---such as optimization that explores a large design space, risk analysis that considers hundreds of different scenarios, or digital twinning that supports in-situ decision making---in practical engineering settings with limited computing resources and turnaround time. We wish to achieve this in an automated manner, without relying on the user expertise, to bring the full potential of simulation-based analysis to not just numerics and domain experts, but to all researchers and practitioners in industry, government, and academia. To this end, we develop model reduction methods that leverage rigorous mathematical theories for PDEs and recent advances in machine learning techniques to accelerate the solution of PDEs by several orders of magnitude with minimal loss of accuracy in many-query and real-time scenarios. In three distinct but tightly coupled topics, we explore reduced-order models that use nonlinear approximation spaces; are provably nonlinear-stable and structure preserving; and provide reliable error estimates and application-specific integration. Our focus is to extend the capability of model reduction from simple "textbook" problems to complex industrial-scale problems. The program seeks to develop novel computational algorithms; develop associated actionable mathematical theories that validate algorithms and inform algorithm development; demonstrate their effectiveness using industrial-scale problems; and transfer the knowledge and capabilities to industry, government, and academia.
The development of advanced model reduction techniques for industrial-scale problems is well aligned with the ongoing effort in the model reduction community and also makes timely contributions to the emerging field of scientific machine learning. This fundamental program will complement ongoing and future applied research projects performed in collaboration with Canadian industry and government organizations. The program will also train students with technical skills in the field of computational science and engineering, a rapidly growing field that combines applied mathematics, computer science, and engineering/science; and professional skills ranging from effective communications to creating and promoting equitable, diverse, and inclusive environments.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
RGPIN-2024-04831
Name nom
Yano, Masayuki
Department département
Institute for Aerospace Studies
Organization id
31
Institution établissement
University of Toronto
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
46000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
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
703
Area of application group en
Transportation systems and services
Area of application group fr
Services et systèmes de transport
Area of application en
Aerospace
Area of application fr
Aérospatial
Research subject code
2108
Research subject group en
Mechanical engineering
Research subject group fr
Génie mécanique
Research subject en
Aerospace, aeronautical and automotive engineering
Research subject fr
Génie aérospatial, aéronautique et automobile
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
Towards model reduction for efficient, reliable, and automated physics-based simulation of industry-scale problems
Keywords
numerical methods; computational fluid dynamics; partial differential equations; reduced order modeling; machine learning; high performance computing
Application summary
With advances in computing hardware and algorithms, high-fidelity simulations based on partial differential equations (PDEs) have become an indispensable tool in the design, analysis, and operation of transportation and energy systems. Continued advances in simulation capabilities are crucial to design next-generation engineering systems that are more efficient, safer, more sustainable, and more economical.
My research program focuses on the development of efficient, reliable, and automated computational methods for PDEs. We wish to provide efficient and reliable PDE solutions to address challenging many-query and real-time tasks---such as optimization that explores a large design space, risk analysis that considers hundreds of different scenarios, or digital twinning that supports in-situ decision making---in practical engineering settings with limited computing resources and turnaround time. We wish to achieve this in an automated manner, without relying on the user expertise, to bring the full potential of simulation-based analysis to not just numerics and domain experts, but to all researchers and practitioners in industry, government, and academia. To this end, we develop model reduction methods that leverage rigorous mathematical theories for PDEs and recent advances in machine learning techniques to accelerate the solution of PDEs by several orders of magnitude with minimal loss of accuracy in many-query and real-time scenarios. In three distinct but tightly coupled topics, we explore reduced-order models that use nonlinear approximation spaces; are provably nonlinear-stable and structure preserving; and provide reliable error estimates and application-specific integration. Our focus is to extend the capability of model reduction from simple "textbook" problems to complex industrial-scale problems. The program seeks to develop novel computational algorithms; develop associated actionable mathematical theories that validate algorithms and inform algorithm development; demonstrate their effectiveness using industrial-scale problems; and transfer the knowledge and capabilities to industry, government, and academia.
The development of advanced model reduction techniques for industrial-scale problems is well aligned with the ongoing effort in the model reduction community and also makes timely contributions to the emerging field of scientific machine learning. This fundamental program will complement ongoing and future applied research projects performed in collaboration with Canadian industry and government organizations. The program will also train students with technical skills in the field of computational science and engineering, a rapidly growing field that combines applied mathematics, computer science, and engineering/science; and professional skills ranging from effective communications to creating and promoting equitable, diverse, and inclusive environments.