Data and code for: Time-Dependent Truck–Drone Cooperative Green Delivery Routing Considering Energy Consumption and Time Windows: A Drone-Aware Adaptive Large Neighborhood Search with Label-Filtering Decoding

Companion data and code of the manuscript Time-Dependent Truck–Drone Cooperative Green Delivery Routing Considering Energy Consumption and Time Windows: A Drone-Aware Adaptive Large Neighborhood Search with Label-Filtering Decoding (Tao Shang, Changzhou University Huaide College). CZ-TDG: a Changzhou (Jiangsu, China) truck–drone green delivery benchmark built from OpenStreetMap — road network, 2,477 candidate stops, five logistics sites and 80 no-fly polygons; 20 instances with 20–200 customers, 10 instances with 8 customers, 6 tuning instances, time-window-scaled variants and 12 Solomon-derived instances; drone energy-model calibration against the published DJI FlyCart 30 data. Code (Python/Numba): time-dependent truck fuel (CMEM) and payload-dependent drone energy models, the label-filtering dynamic-programming decoder, DA-ALNS and the baselines S-ALNS, Q-ALNS, HGA-D and PyVRP+DP, exact enumeration, decoder test, experiment job lists, scenario analyses, table and figure scripts. Run records: HGA-D and PyVRP+DP runs of the main and n = 8 experiments, HGA-D tuning runs, enumerated optima of the n = 8 instances, decoder-test records, best-known costs and the CZ100-GLP best solution (see README.md for what is and is not included). The package is one 7z archive (CZ-TDG.7z.001) with an MD5 manifest; see README.md and HOW_TO_REASSEMBLE.txt. OpenStreetMap-derived data: © OpenStreetMap contributors, ODbL 1.0; other data and results: CC BY 4.0; code: MIT.