Researchers at Uppsala University have developed an AI model that accurately grasps how electric vehicle batteries age and deteriorate over time.
This is because electric vehicles are wearing fast batteries, which slows down the electrification of the transportation sector.
The newer models may increase the longer life and safety of these batteries, allowing for faster electrical transitions.
Professor Daniel Brandel, who led the research and oversees the Ångström Advanced Battery Center at Uppsala University, explained:
“If you simply stop looking at it as a black box that is expected to provide power and instead get a detailed image of the process, you can manage them and keep them in their state for longer.”
EV batteries are the first component of age
It is not uncommon for an electric vehicle battery to become the first component of an age vehicle. This is a huge waste of resources today and is hindering the transformation of the transportation sector.
Over time, lithium-ion batteries, the most common type used in EVs, gradually lose their ability to hold a full charge. There are several factors that affect this degradation, such as charge/discharge cycles, high temperatures, fast charging, and depth of discharge.
Normally, EV batteries can lose about 2-3% of their annual capacity, but the rates vary. After 8-10 years of regular use, the EV battery can retain 70-80% of its original capacity. This often allows the vehicle to function, but may reduce driving range.
To address this issue, the automotive industry is often developing software based on AI to optimize battery management and control.
Models created by Uppsala University can increase the robustness of battery health forecasts by up to 70%.
Battery cycle life mapping
The several years of battery testing is behind a study conducted in collaboration with Aalborg University in Denmark.
A database was built by collecting data from a large number of very short charging segments. It was then combined with a detailed model of all the different chemical processes that take place within electric vehicle batteries.
“Overall, this gives us a very accurate picture of the various chemical reactions that the battery produces, but we also draw very accurate pictures of how it ages during use,” commented Wendy Guo, who conducted the study.
Reduce the need for confidential vehicle data
This discovery can also address the safety of electric vehicle batteries. Safety issues that can occur with batteries are often due to design flaws or side reactions, and can also be predicted by studying battery charging and discharging data.
Brandell concluded: “The fact that we only use short charging segments is probably an added benefit. Battery data for electric vehicles is sensitive to the industry and to users from an anonymization perspective.
“This study shows how far you can get without the need for a complete data set.”
Source link
