The ongoing transformation of the mobility sector to- wards more environmentally friendly transport systems is leading to increasing market penetration of electrified mobility solutions. As part of these solutions, personal micromobility represents a steadily growing market. The need for fast charging options results in increased requirements for thermal management and state estimation algorithms to operate battery storage systems within acceptable operating ranges. Within the scope of the Safeμmob project, the potential of impedance-based temperature determination at cell level in a pedelec is to be investigated. This is a commercially available system that has been extended by a specially developed battery management system to test model-based algorithms in a real system. High-resolution sensor technology in conjunction with a data logger enable not only the testing of real-time applications but also the use of the pedelec to analyse the drive behaviour
The pedelec system, as shown in Figure 1, assists the rider up to a speed of 25km/h, with a maximum power of 250 W. Depending on the controller implementation, the set assistance level and the basic system characteristics, low-frequency (1 − 5 Hz) current fluctuations occur in the phases with electrical assistance, which are due to the cadence. Amplitudes of up to 300 mA provide a sufficient excitation signal for the online impedance determination in the investigated system.
Temperature-dependent characteristic maps result from the systematic characterization of the battery cell used by electrochemical impedance spectroscopy.
For the reconstruction of the average cell temperatures, the temperature dependence is modeled via the Arrhenius law at 2.4 Hz.
The proof of concept shows that the occurring loads in acceleration and climbing phases are dominated by the cadence, so that an impedance determination for 1.5 Hz < f < 3 Hz leads to plausible results. The reconstructed temperatures show plausible curves compared to the integrated NTC. Initial differences in cell impedances can be attributed to construction-related resistances and different ageing states. In further work, in addition to the influence of ageing, the current dependence in the system must also be investigated and the temperature determination must be validated. The conclusion on the average cell temperature requires the integration of model-based estimators and a quality determination of the measurement data.