Published August 6, 2026 | Version paper-submission-v2.0

External Validation Before Model Complexity: Reproducibility Package for Recursive Lithium-Ion Battery RUL Forecasting

Authors/Creators

  • 1. College of Electrical and Information Engineering, Heiljiang Bayi Agricultural University, Daqing, China

Description

This record archives the reproducibility materials associated with the manuscript “External Validation Before Model Complexity: A Five-Gate Evaluation of Recursive Lithium-Ion Battery RUL Forecasts.”

The archive contains: (1) the HUST post-confirmation v3 analysis package, including converted trajectories, gate metadata, protocol stress-test outputs, exploratory timing-utility sensitivity results, configuration files, execution logs, analysis code, and independent recomputation audits; and (2) the strengthened Journal of Energy Storage submission materials, including the manuscript, supplementary material, graphical abstract, highlights, cover letter, and revision documentation.

The frozen HUST H1/H2 one-shot confirmation remains unchanged. The post-confirmation analyses evaluate five controlled protocol violations and exploratory utility contracts; they do not relabel, repair, or replace the original confirmatory result. HUST is classified as untouched confirmatory wrong-direction transfer, the XJTU Batch 4 cycle-190 result as amended conditional benefit, and the registered BIT primary comparison as non-estimable.

Exact file identity is documented through accompanying SHA-256 checksum files. Original third-party battery datasets are cited in the manuscript and remain subject to their original licenses.

Software and analysis scripts are released under the MIT License. Manuscript materials, documentation, figures, tables, and derived result summaries are released under CC BY 4.0. Third-party source datasets remain under their original licenses.

Files

MSTT_RUL_JES_STRENGTHENED_V10_Submission_Package.zip

Files (15.2 MB)

Name Size Download all
md5:d5d78cd970d9274b0f10bd1b77a45090
2.8 MB Preview Download
md5:084b74b623bf381a99b1a66c841a0a19
129 Bytes Download
md5:c0088fae50cb816a0d4f4ab343db8c98
7.6 MB Preview Download
md5:b3ae45452e37cebd13743f78c3d7e18d
112 Bytes Download
md5:973acf9d17643be4b80f8353c8c6c5c6
4.8 MB Download
md5:00cc1de9933ce77864d3470ea55a73d1
118 Bytes Download

Additional details

Related works

Is new version of
Dataset: 10.5281/zenodo.21758262 (DOI)