DiffLiB: High-fidelity differentiable modeling of lithium-ion batteries and efficient gradient-based parameter identification
Weipeng Xu, Kaiqi Yang, Yuzhi Zhang, Shichao Sun, Sheng Mao, Tianju Xue
TL;DR
DiffLiB tackles the challenge of inverse parameter identification for the full-order Doyle–Fuller–Newman (DFN) battery model by introducing a differentiable finite-element framework built on JAX-JAX-FEM. It enables end-to-end gradient-based optimization by combining an adjoint-based implicit differentiation method with automatic differentiation, and employs a multi-scale sequential solution to manage the large macro/micro DOF coupling. Across 2D and 3D forward benchmarks and parameter-identification studies, DiffLiB achieves COMSOL-grade forward accuracy (RMSE < 2 mV) while delivering substantial computational gains for inverse problems (up to 96% fewer forward evaluations and 72% shorter compute time). The results demonstrate DiffLiB’s potential for efficient, high-fidelity calibration and design optimization of high-dimensional LIB systems with realistic 3D geometries.
Abstract
The physics-based Doyle-Fuller-Newman (DFN) model, widely adopted for its precise electrochemical modeling, stands out among various simulation models of lithium-ion batteries (LIBs). Although the DFN model is powerful in forward predictive analysis, the inverse identification of its model parameters has remained a long-standing challenge. The numerous unknown parameters associated with the nonlinear, time-dependent, and multi-scale DFN model are extremely difficult to be determined accurately and efficiently, hindering the practical use of such battery simulation models in industrial applications. To tackle this challenge, we introduce DiffLiB, a high-fidelity finite-element-based LIB simulation framework, equipped with advanced differentiable programming techniques so that efficient gradient-based inverse parameter identification is enabled. Customized automatic differentiation rules are defined by identifying the VJP (vector-Jacobian product) structure in the chain rule and implemented using adjoint-based implicit differentiation methods. Four numerical examples, including both 2D and 3D forward predictions and inverse parameter identification, are presented to validate the accuracy and computational efficiency of DiffLiB. Benchmarking against COMSOL demonstrates excellent agreement in forward predictions, with terminal voltage discrepancies maintaining a root-mean-square error (RMSE) below 2 mV across all test conditions. In parameter identification tasks using experimentally measured voltage data, the proposed gradient-based optimization scheme achieves superior computational performance, with 96% fewer forward predictions and 72% less computational time compared with gradient-free approaches. These results demonstrate that DiffLiB is a versatile and powerful computational framework for the development of advanced LIBs.