Fusion of measurements by existing nodes in an on-board power supply system for condition monitoring with focus on the battery

Abstract:
As the transition from fossil fueled to electrically powered passenger vehicles takes place, the battery becomes an even more prominent system in the car. From a variety of exiting technologies, the lithium ion battery is with good prospects to become the dominating technology. Despite the good properties of lithium ion batteries, safety is a main drawback and a certain range for the state of charge (SoC) must be assured. To achieve that, precise SoC estimation measures are necessary. In vehicles, every component serves a distinct function and is rationalized whenever possible. This work focuses on extracting impedance based features utilizing a simplified board net structure as excitation source. In that scenario, the Taylor Fourier transform (TFT) is used to extract features, which can be used for battery state detection. Challenges and opportunities of the method are discussed. From the perspective of simulations, the deviations of the impedance can be used to improve the precision of SoC estimations, e.g. in an application with machine learning methods.