CHALLENGE

Currently states indicators are estimated using only a limited amount of data, such as voltage and current measurements.

Intelligent sensor technologies can directly measure physical chemical parameters, which provides a significant operational advantage.

For example, fast charging strategies can be improved and the usable capacity increased because safety margins do not have to be excessively large.

OBJECTIVE

The main goal is to select a suitable sensor technology depending on a given system and to collect the data generated by operating the sensor in situ.

Depending on the operating strategy of the sub-system (Battery, FCs, e-motor, power electronics), an indicator for a state to be defined is to be selected and calculated.

This indicator can provide information to the operating strategy in the Component Management System (CMS) during operation, thus enabling more optimal operation.

APPROACH

The plan is to train a model with preliminary tests, which cleans the measurement of the sensor from the current side effects based on all available measurement data. This will be led by the RWTH Aachen University based on years of experience in the field of AI.

Additional knowhow will be made available to RWTH by secondments from other partners. To this end, the measurement setup and the hardware required for it should be in operation by the time of secondment.