Quiet-State-Anchored Simulation-to-Reality Adaptation for Arcing High-Impedance Fault Detection: Code and Per-Run Results

Simulation generators (MATLAB/Simulink), training, adaptation and evaluation code (Python/PyTorch), per-run results, trained detector weights, generated tables and figure scripts for the article "Quiet-State-Anchored Simulation-to-Reality Adaptation for Arcing High-Impedance Fault Detection" by Z. Tian, S. Ke, H. Song and R. Xue (Energies, 2026, accepted). qsa-hif-v1.0-code.zip contains the source code; qsa-hif-v1.0-results.zip contains the results.json and scores.npz files of the 135 principal, ablation and threshold-control runs, the revision runs, the weights of the control-retrain detectors and the generated tables. The physical data sets (VeHIF, the covered-conductor campaign and WHITED) are public and are not redistributed. README.md lists the data sources, the software environment and the mapping from the tables and figures of the paper to the scripts. Code: MIT License; results, weights and tables: CC BY 4.0.