Identification of Battery Aging Based on the Dynamic Error of the Chen-Lee Chaotic System and Gramian Angular Difference Field Applied for Lithium Battery Relaxation Voltage

This paper suggests a characteristic analysis approach for lithium - iron - phosphate battery relaxation voltage. It uses the Chen - Lee chaotic system dynamic error equation to transform the time series of normal and test relaxation voltage waveforms into one - dimensional dynamic error values. Then, the Gramian Angular Difference Field (GADF) method extracts key features reflecting time frequency and dynamic structural differences via inner product and angle transformation. This method retains key info, reduces dimensions, and enables effective feature extraction. Moreover, the paper explores converting GASF's time - correlation info into 2D images. These images are analyzed by the structural similarity index (SSIM), which gauges similarity by extracting image texture and color features. This helps judge whether lithium - iron - phosphate battery aging needs immediate maintenance or monitoring. The study verified the method using ANR26650M1-B lithium - iron - phosphate batteries. Results show that combining the Chen - Lee chaotic system, GADF - based feature extraction, and SSIM similarity evaluation can effectively identify battery aging statuses, enable predictive diagnosis, and inform decisions on battery replacement. The method shows aging warning potential.