Simulation and verification data for task-persistent hybrid residual-memory quadrotor tracking
Description
This collection contains the frozen simulation programs, trajectory-level outcome tables, paired statistical summaries, public-flight residual-prediction replay summaries, figure files, and deterministic certificate-verification artifacts supporting the accompanying quadrotor-control manuscript. The primary and stress studies use software models; the RotorPy transfer study uses an independently implemented software plant. NanoBench replay keeps the recorded trajectory and commands fixed and evaluates residual predictors; it is not a flight or hardware-in-the-loop test of the proposed controller. The archive preserves the existing 66-file revision supplement byte for byte and adds repository documentation, a column-level data dictionary, file descriptions, provenance, and checksums. Third-party NanoBench raw logs are not redistributed; their pinned public source and retrieval commands are documented. Complete sample-level time histories from the 960 closed-loop method-run records are not included; the scripts and frozen seeds support regeneration. The added hover verification concerns matched nominal local dynamics, not a certified attraction region for the whole uncertainty rectangle. The author's original data and documentation are under CC BY 4.0; the author's original code is under MIT.
Files
Steps to reproduce
Download and extract TPHRM_Research_Data_Code_Release_v1.zip. From TPHRM_reproducibility, create a compatible Python 3.9 environment, install requirements-core.txt, and run python run_revision_verification.py. This checks the archived outcomes and deterministic identities; it does not regenerate all 960 closed-loop method-run records. Full ensemble and optional RotorPy commands, frozen seeds, and the public NanoBench retrieval/replay procedure are in README.md and REPRODUCIBILITY_SPECIFICATION.md. Third-party raw flight logs are obtained separately from the pinned upstream source in PROVENANCE.md.