AI-based modular battery systems for commercial and grid applications

The KI-M-Bat project focuses on developing and testing a modular storage concept and a universal AI-based software control system for battery storage systems on an industrial and commercial scale. The concept enables efficient and lifetime-optimized operation of second-life batteries. A digital twin model of the battery system will be built into the software, and a control strategy will be tested using machine learning. The algorithms will be validated in a hardware demonstrator and published as open-source code.

TUM-EES will contribute its expertise in the areas of data-based battery modeling, aging simulation and experience in the operation of storage systems. Based on the open-source tools developed at EES, an open-source digital twin model of the battery modules will be developed in the project. Parameterization will be based on historical field data and directly linked to the hardware demonstrators.

The project pursues the following goals:

• Parameterization of battery models from field data and online state estimation at the module level.
• Enabling the operation of second-life battery systems with different cell aging states and cell chemistries.
• Significant extension of battery service life through AI-supported aging-sensitive control of the battery modules.

In addition to the Chair of Electrical Energy Storage Technology at TUM, the project also involves:

• STABL Energy GmbH
• Fenecon GmbH
• Kempten University of Applied Sciences
• The TUM Center of Combined Smart Energy Systems (CoSES)

More information on the project is available at https://stabl.com/forschung/