Battery SoH Estimation Through Non-Destructive Methods

The growing proliferation of electric and hybrid vehicles in recent years has led to a rapid and significant acceleration in the evolution of energy storage technologies. This, in turn, has generated an increasingly pressing need to monitor the health status of battery cells with precision, consistency, and standardization, especially now that the second-hand market for such vehicles is steadily expanding, along with the amount and variety of data that can be acquired from batteries used in diverse operating conditions. This thesis stems from the need to develop a standardized and easily reproducible diagnostic system for assessing battery health in the automotive sector. The focus of the study is on the Electrochemical Impedance Spectroscopy (EIS) procedure. The initial goal was to infer the State of Health (SoH) of the batteries through impedance measurements at specific frequencies. However, as the study progressed, it became increasingly evident that this methodology holds a broader potential, capable of providing a more comprehensive representation of the battery health state. The analysis begins with a critical review of the existing non-destructive diagnostic techniques for batteries, explaining the methodological choices adopted. Once the diagnostic approach was selected and the theoretical analysis completed, the experimental procedure is presented in detail, it consisted in controlled laboratory measurements conducted on cylindrical batteries 18650 typically used in domestic applications, allowing the collection of data that were both repeatable and interpretable. The development of an equivalent circuit model, together with a dedicated code capable of representing the data in a useful and meaningful way, enabled comparison with real-world data and the extraction of valuable insights from them. Finally, the thesis discusses the prospects for industrialization, highlighting the remaining challenges in terms of miniaturization, signal robustness in noisy environments, and the inherent difficulty of conducting a large-scale data collection campaign.