The application of pulse response analysis method in lithium-ion battery modeling and state estimation

Battery management depends on accurate battery models. However, comprehensive battery models are often complicated and computationally demanding, limiting their practical usage, Researchers aim to simplify and improve models for better system efficiency. This paper introduces a pulse response (PR) analysis method to describe battery polarization characteristics . By combining PR analysis, convolution theory, Kalman algorithm, and regression algorithm, we propose a precise calculation method for the battery's pulse response function and establish a simplified battery model structure. Under dynamic conditions, the simulated terminal voltage error using the pulse response model is <0.4 %, showcasing its accuracy. Additionally, the State of Charge (SOC) estimation error remains below 0.5 %, further highlighting the model's reliability. Furthermore, the paper introduces a rapid State of Health (SOH) estimation approach using pulse response functions and time coefficient (TC) methodology. This method achieves precise SOH estimation with an error of <0.2 % through linear regression based on a 1 s pulse and 5 s relaxation. Our method effectively avoids the subjective biases that conventional methods may encounter in parameter identification and feature extraction, providing a more objective basis for battery simulation and management. This makes our method highly suitable for engineering applications and contributes to enhancing battery management performance. • High-precision method for obtaining the pulse response function • Describing battery polarization using pulse response functions • High accuracy in voltage simulation and state estimation • Estimate battery SOH with high accuracy using 5 s relaxation data