An EV-Scale Demonstration of In-Situ Battery Electrochemical Impedance Spectroscopy and BMS-Limited Pack Performance Analysis

State-of-power (SOP) and state-of-charge (SOC) estimation inaccuracy manifests throughout the electric vehicle (EV) battery lifetime due to the nonlinear degradation trajectory that is unique to each use-case. The inaccuracy leads to premature termination of charge and discharge operations, which results in the virtual loss of battery performance. In-situ electrochemical impedance spectroscopy (EIS) has been proposed to provide real-time battery impedance measurements, which can theoretically improve the SOP/SOC estimation accuracy. In this work, an EV-scale in-situ EIS system is demonstrated experimentally, from impedance measurement to equivalent circuit model (ECM) extraction. The pack-level discharge energy versus SOP/SOC estimation errors is also simulated, which quantifies the virtual loss of battery performance. The measured-impedance error tolerance of three common ECM extraction techniques is then examined through a randomized set of simulated EIS tests with injected measurement noise. Finally, in-situ impedance measurement is performed on 20 battery submodules using an electric pickup truck EIS system. The ECM-predicted voltage output is compared against measured driving data. Experimental results indicate approximately 8%-9% improvement in SOP estimation accuracy at 3 degrees C and 10 degrees C between the in-situ extracted models and a representative baseline.

Automated summary

The inaccuracy of state-of-power and state-of-charge estimation in electric vehicle batteries leads to virtual loss of battery performance. An in-situ electrochemical impedance spectroscopy system is demonstrated to improve estimation accuracy. The impact of estimation errors on battery performance is quantified through simulations, and the accuracy of three common extraction techniques is examined. Experimental results show an improvement of approximately 8%-9% in SOP estimation accuracy compared to a baseline model.