An inductive extension of the physics-based model for lithium-ion batteries under fast and non-periodic current excitation

ABSTRACT Physics-based pseudo-two-dimensional (P2D) models are widely used to describe lithium-ion batteries; however, their standard terminal-voltage formulation is mainly valid under slow or moderately varying current conditions. Under fast and non-periodic current excitation, such as pulsed or step-like inputs, measured voltage responses often exhibit inductive features that cannot be captured by conventional P2D models. Such fast and non-periodic current profiles are commonly encountered in practical battery management systems (BMS) and power electronics applications. In this work, the classic P2D model is extended by introducing a lumped inductive term into the terminal-voltage formulation. This extension enables the model to capture fast transient voltage dynamics and the associated high-frequency impedance behavior. The inductive term represents the combined effects of current collectors, internal conductive pathways, and external connections, which are inherently present in practical battery systems and become non-negligible under fast current variations, while preserving the original electrochemical description. The extended model is applied to simulate voltage responses under non-harmonic current excitations, representing non-periodic current profiles encountered in realistic operating conditions. The results show that inductance strongly affects the transient response, leading to voltage overshoot, positive phase angles, and loop-shaped voltage–current trajectories. These features are not reproduced by the conventional model without inductance. The simulated responses remain consistent with experimentally measured impedance characteristics in the high-frequency range. Overall, this work highlights the importance of explicitly accounting for inductive effects in physics-based battery modeling under fast dynamics. The proposed formulation provides a simple and measurement-consistent extension of the P2D framework, improving the description of high-frequency transient behavior and reducing the risk of non-physical parameter compensation during model calibration.