CT Detection Method for Lithium-ion Battery Overhang Region Based on Sampling Point Fitting

Le résumé fourni par la source

Abstract Accurate detection of the overhang (OH) region is critical to the safety and energy density of lithium-ion batteries. Conventional X-ray CT detection relies on four-corner sequential scanning, which suffers from low efficiency, insufficient random error suppression, and poor outlier robustness, making it hard to satisfy industrial high-speed online inspection. To address these issues, this paper proposes a parametric detection method for lithium-ion battery OH regions based on noisy sampling point fitting. The core innovation is abandoning the traditional four-corner scanning and achieving high-precision detection using only double opposite-corner sampling. A geometric model is established with the center coordinates and tilt angle of the positive electrode sheet as optimization targets, transforming OH measurement into a nonlinear least-squares problem. A hybrid PCA+RANSAC strategy is used for robust initial value estimation to resist noise and outliers, followed by high-precision iterative optimization via the Gauss-Newton method. Simulation and real CT experiments show that the proposed method improves detection accuracy by 56.8% and efficiency by 49.8% with 100% convergence rate. It exhibits strong anti-interference ability and good generalization for stacked and wound batteries. This method breaks the limitation of conventional full fourcorner scanning and provides a feasible solution for online, high-speed, high-precision nondestructive detection of lithium-ion battery OH dimensions.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.