Reliable Cell Behaviour Predictions for Real Life Conditions Get Started...
Read MoreFrom Inconsistency to Integrity: How Physics Informed Machine Learning Tackles EV Battery Challenges
Vineet Dravid
Founder, CEO oorja
From Inconsistency to Integrity: How Physics Informed Machine Learning Tackles EV Battery Challenges
Introduction
As EVs become mainstream and production volumes increase, car companies face new challenges, particularly in determining cell quality. Typically, vehicle production lines last for several years. Cells over batches that purportedly have the same part number show significant variations, which can lead to safety and performance issues. Recent data has suggested that the expected variation in capacity and impedance values between batches is around 5% and can go as high as 20% in cases where manufacturing protocols are not stringent. This can significantly compromise the pack design that has been optimized for a particular batch.
These cell variations over batches pose a challenging question: How does a company know the new batch is safe to integrate into the existing design?
oorja has a solution for this problem in the form of THERMAI, a physics-informed ML-based tool that can help engineers take quick decisions on the safety integrity of a new batch without performing extensive thermal safety tests, which are hazardous, expensive and time-consuming.
Workflow
The workflow uses physics-based insights to create a machine learning (ML)-based digital twin that accurately maps the base batch and pack design. Figures 1a and 1b show physical insights extracted from the performed experiments.
Minimal and quick cycler experimentation on the new batch is then used to run the trained model and predict the new batch’s relative safety. Figure 2a illustrates a cell from the new batch exhibiting a TR index of 62%, indicating a 24% higher risk of thermal runaway compared to the original batch.
Using advanced algorithms, an estimated 3D TR propagation model is also created, which assures the pack integrity for the new batch of cells.
To explore how you can use oorja to simulate accelerated testing, contact us at info@oorja.energy
Latest Blog Posts
From Inconsistency to Integrity: How Physics Informed Machine Learning Tackles EV Battery Challenges
From Inconsistency to Integrity: How Physics Informed Machine Learning Tackles...
Read MoreDeciphering Charging Dynamics: Part 1 Why Does It Take So Long To Charge My Battery?
Deciphering Charging Dynamics: Part 1 Why Does It Take So...
Read More