Mission-Oriented Four-Dimensional UAV Trajectory Planning with Reinforcement Learning
Year: 2026
Type: dataset
Abstract: Supporting simulation data for “Mission-Oriented Four-Dimensional UAV Trajectory Planning with Reinforcement Learning”. The dataset includes per-task evaluation records for four planning methods on 200 shared test missions, CPA-feature ablation records, training-reward scalars, representative four-dimensional trajectories, the fixed mission bank, and static simulation geometry. Field definitions, data provenance, checkpoint identities, common-success task membership, and scripts for recomputing the numerical summaries are provided. The no-CPA ablation uses a fallback checkpoint for training seed 2026 and final checkpoints for seeds 2027 and 2028, as documented in the package.
Author: ZHANG, CHUAN HUA
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7221517688