Combining machine learning and electrochemical models

(Image courtesy of Eatron)

Eatron Technologies and About:Energy are developing a decision-engine for battery management systems (BMS) that combines machine learning and electrochemical models, writes Nick Flaherty.

The aiMAGINE project aims to use machine learning frameworks in the BMS to extend the life of a battery pack in electric vehicles and scooters.

Current BMS rely on simple, empirical methods that sacrifice accuracy in return for reduced computational effort. However conventional AI methods remain challenging to integrate within the BMS due to their complexity, demanding training process, and the need for large volumes of input data.

The aiMAGINE project combines About:Energy’s electrochemical battery models that achieve rapid and accurate calibration with Eatron’s edge and AI-powered cloud platform. This should provide more accurate assessments of state-of-charge (SoC), state-of-health (SoH) and (patented) remaining useful life predictions.

The AI complements the electrochemical models, enhancing predictions by accounting for complex physical behaviours that cannot be modelled. This will allow the AI-powered decision engine (AI-DE) to provide highly accurate operational parameters to the BMS, significantly increasing battery pack longevity and simplifying integration.

“Implementing our novel AI-powered intelligent battery software layer with this revolutionary AI-DE can extend a battery pack’s first life by up to 20%,” said Dr Umut Genc, CEO of Eatron.