A battery modeling framework driven by the distribution of relaxation times reduced terminal voltage prediction root mean square error to 11.97 millivolts under UDDS operating conditions. This performance marks a 31.16 percent improvement in prediction accuracy over conventional models, while also decreasing the average frequency-domain fitting root mean square error by 19.7 percent.
High-precision battery modeling is necessary for dependable state estimation in battery management systems. However, conventional fractional-order equivalent circuit models generally employ empirical strategies to determine topological order and lack consistency verification between the frequency and time domains.
To resolve these constraints, the proposed framework applies distribution of relaxation times technology to decouple multi-scale polarization kinetics within batteries, allowing model order to be determined on a physical basis. For robust parameter identification, the methodology incorporates a physics-constrained two-stage optimization algorithm combining particle swarm optimization and sequential quadratic programming.
The researchers also evaluated parameter sensitivity across diverse testing environments. Their evaluation indicated that pairing an appropriate excitation amplitude with sufficient rest time establishes an optimal experimental configuration for high-precision parameter extraction in fractional-order battery modeling.
The researchers emphasize that this framework provides a reliable technical reference for engineering battery testing and high-robustness battery fractional-order equivalent circuit modeling.
Subject of Research: Distribution of relaxation times-driven battery modeling and parameter sensitivity evaluation
Article Title: Distribution of relaxation times-driven battery modeling: Multidimensional comparison and parameter sensitivity evaluation
Article References: Qin, A., Wei, X., Ma, T., Wu, L., Lyu, Z., & Ma, D. (2026). Distribution of relaxation times-driven battery modeling: Multidimensional comparison and parameter sensitivity evaluation. Ionics. https://doi.org/10.1007/s11581-026-07650-5
Image Credits: AI Generated
DOI: 10.1007/s11581-026-07650-5
Keywords: distribution of relaxation times, battery modeling, fractional-order equivalent circuit models, parameter sensitivity evaluation, particle swarm optimization, sequential quadratic programming, PSO-SQP algorithm, polarization kinetics, model order determination, frequency-domain fitting, root mean square error, UDDS operating conditions, terminal voltage prediction, battery management systems, excitation amplitude, rest time, parameter identification, cross-domain consistency
News Source: Faith Mcneil. (November 18, 2026). Relaxation Time Battery Model Cuts Voltage Prediction Error by 31 Percent. Scienmag.



