Data supporting “Health–Capability Ambiguity in Degraded Aircraft: Task-Relevant Residual Capability Representation for Mission-Entry Decisions”

Description

This dataset provides the processed evidence and frozen protocol metadata supporting the manuscript “Health–Capability Ambiguity in Degraded Aircraft: Task-Relevant Residual Capability Representation for Mission-Entry Decisions.” It includes the 60-key exact-grid demand definition, degradation-state metadata, capability-certification records, controlled 200-mission comparison data, capability-estimation sensitivity results, independent 300-case validation data, scope-boundary summaries, protocol metadata, and numerical source data for the reported figures and tables. The package is designed to reproduce the principal numerical results reported in the manuscript. It does not include the full simulation-development repository, full 120-Hz trajectory histories, the complete 15,849 transition-level raw dataset, or future transition-aware research assets.

Steps to reproduce

The released dataset contains the processed evidence and frozen protocol metadata required to reproduce the principal numerical results reported in the associated manuscript. Use 01_demand_grid/demand_keys_60.csv and degradation_states.csv to reconstruct the finite exact-grid demand domain and evaluated degradation states. Each degradation state contains 60 task-demand keys: 3 steady-speed, 9 coordinated-turn, 21 climb/descent, and 27 coupled turn–climb keys. Use the files in 02_capability_certification/ to reproduce the certified capability memberships, health–capability ambiguity rates, and pairwise Jaccard similarities reported in the manuscript. Use 03_controlled_comparison/ to reproduce the 200-mission comparison among B0, B1, B2, B3, A1, and the Proposed representation, including false-feasible, completion, and accepted-plan completion statistics. Use 04_sensitivity/ to reproduce the capability-estimation perturbation results. Use 05_independent_validation/validation_300.csv to reconstruct the confusion matrices and calculate precision, recall, and balanced accuracy for the 300-case independent validation. Use the aggregate files in 06_scope_boundary/ to reproduce the held-out route, paired-random route, and transition-level scope-boundary statistics. The numerical source data for Figures 3–6 and Tables 3–4 are provided in 08_figure_table_source/. Qualification thresholds and execution settings are documented in 07_protocol/. Field definitions and units are provided in DATA_DICTIONARY.md. This release does not contain the complete simulation-development repository, full 120-Hz trajectory histories, or the complete 15,849 transition-level raw dataset.

Institutions

• North China Institute of Aerospace Engineering

  Hebei, Langfang

Categories

Licence