McMaster University
Practical implementation of electrochemical impedance spectroscopy in battery characterization and management
Abstract
dc:description.abstractElectrochemical impedance spectroscopy (EIS) offers rich, frequency-resolved information about the internal state of lithium-ion batteries, but its practical use onboard an operating battery system is limited by the relaxation effect, which requires the battery to rest for hours before a stable impedance can be measured. This thesis addresses this limitation through three complementary bodies of work. The first part experimentally demonstrates that the previous-cycle load condi tions, in addition to the rest period, govern the apparent impedance. The concept of the state-of-ionic-diffusion (SOID) is introduced to unify their effects. Repeatable impedance measurements are shown to be achievable after a rest of 5 minutes or less when the SOID is controlled. The second part proposes a predictor-corrector framework, combining two models through an unscented Kalman filter (UKF), to mitigate relaxation-induced measurement uncertainty in state-of-health (SOH) estimation. Validated on open-source datasets of 51 cells, the framework achieves an SOH estimation error of 1.4–1.7% and generalizes to previously unseen cell chemistries. Applied to a two-year field dataset from 20 electric vehicles (EVs), the same framework yields more accurate and robust estimates than conventional approaches, particularly under noisy conditions. The third part derives an analytical expression for the sensitivity of the impedance of a single-particle model with double-layer capacitor (SPM-DL) to its parameters and characterizes the sensitivity function behaviours across the frequency range. It not only builds towards a better understanding of the impedance relaxation process, but also supports applications in parameter identification, uncertainty quantification and fault detection. Two earlier studies preceded the core of the thesis. The first examines the accuracy of equivalent-circuit model (ECM) parameters when fitted to different experimental load profiles, highlighting the dependency of time-domain analysis on the load profile. The second demonstrates the potential of EIS in module-level applications. Together, these studies characterize, mitigate, and begin to explain the relaxation effect, advancing the practical implementation of EIS as an onboard diagnostic tool for battery systems.
Degree
thesis:*- Department dc:contributor.department
- Mechanical Engineering
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Zhang, Wenlin
- Advisor dc:contributor.advisor
-
- Habibi, Saeid
Rights
- Language dc:language.iso
- en
Identifiers
dc:identifier.*- Identifier URI
- https://doi.org/10.71548/25417