Project Details
Description
Abstract:
Lithium-ion batteries are essential for achieving a low-carbon future, powering applications from consumer electronics to electric vehicles and renewable energy systems. However, their performance and lifetime are influenced by a combination of factors, including chemistry, design, and operating conditions. The complex and nonlinear interactions among these factors pose significant challenges for effective battery management. Digital twins offer a promising solution to address these problems, integrating real-time sensor data with advanced computational models to create a dynamic digital representation of the battery system. This approach facilitates predictive insights and data-driven decision-making for optimal battery operation.
This PhD project aims to develop a multi-physics and AI accelerated battery digital twin to advance the state-of-the-art in battery management. The main contribution lies in combining a computationally efficient physics-based model - capturing key internal states such as lithium-ion concentration, overpotential, and temperature - with AI techniques for real-time monitoring and enhanced prediction. The research objectives include (1) designing scalable multi-physics models to capture coupled electrochemical, thermal, and mechanical dynamics, (2) customizing the model for commercial BYD 4680 battery cells to achieve accurate state estimation, (3) integrating AI algorithms with multi-physics models for accelerated computation, and (4) validating the digital twin framework with experimental data on AI dedicated hardware systems. The expected outcomes of this project include a comprehensive and computationally efficient model, accurate state predictions, improved battery performance, extended lifespan, and enhanced safety.
Funding: VillumFoundation founded project: Smart Battery
Lithium-ion batteries are essential for achieving a low-carbon future, powering applications from consumer electronics to electric vehicles and renewable energy systems. However, their performance and lifetime are influenced by a combination of factors, including chemistry, design, and operating conditions. The complex and nonlinear interactions among these factors pose significant challenges for effective battery management. Digital twins offer a promising solution to address these problems, integrating real-time sensor data with advanced computational models to create a dynamic digital representation of the battery system. This approach facilitates predictive insights and data-driven decision-making for optimal battery operation.
This PhD project aims to develop a multi-physics and AI accelerated battery digital twin to advance the state-of-the-art in battery management. The main contribution lies in combining a computationally efficient physics-based model - capturing key internal states such as lithium-ion concentration, overpotential, and temperature - with AI techniques for real-time monitoring and enhanced prediction. The research objectives include (1) designing scalable multi-physics models to capture coupled electrochemical, thermal, and mechanical dynamics, (2) customizing the model for commercial BYD 4680 battery cells to achieve accurate state estimation, (3) integrating AI algorithms with multi-physics models for accelerated computation, and (4) validating the digital twin framework with experimental data on AI dedicated hardware systems. The expected outcomes of this project include a comprehensive and computationally efficient model, accurate state predictions, improved battery performance, extended lifespan, and enhanced safety.
Funding: VillumFoundation founded project: Smart Battery
| Status | Active |
|---|---|
| Effective start/end date | 01/12/2024 → 30/11/2027 |
Fingerprint
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.