NeuroBatt

In order to ensure the best performance and lifetime of a battery system, the included battery management system (BMS) should enable the monitoring of the state of charge (SOC) and of the state of health (SOH) precisely in an early state. The idea behind the project Neurobatt is to acquire a full spectrum of information from the battery pack and use it for state of health prediction based on machine learning algorithms. Due to the fact that temperature plays an extremely important role in cell aging and failure, the temperature of each cell is performed by glass fiber sensors as well as the measurement of voltage and current during operation. By modulating the charge current with a multi–sine probe signal, it is possible to obtain the continuous time-varying impedance of each cell of a battery in use. This is called Dynamic Electrochemical Impedance Spectroscopy (DEIS).

This project is conducted by a consortium of independent companies and research centers to bring together the expertise to develop all the necessary parts of the project, namely: hardware and electronics development, cell and electrodes manufacturing, materials characterization, and artificial intelligence aided data analysis.

The Modelling and Simulation group of Fraunhofer IFAM led by Prof. La Mantia works on the development of an experimental set-up for the continuous acquisition of multi-frequency modulated voltage and current signals as well as data processing and computation.

Another task of the project is the evaluation of the enhancement of the prediction accuracy from the utilization of a three-electrode cell format with reference electrode for correct potential estimation of the electrodes.

After the successful demonstration of the time-varying impedance acquisition for a laboratory-scale cell, the method is under implementation in a 12 cell battery pack and integration in the BMS microcontroller.