无人机群智行为数据集

无人机群智行为数据集面向群智涌现行为的表示与机器学习识别技术的研究，虚实结合强化学习的训练和验证需求建设，基于无人机虚实结合行为识别、训练和验证工具UXVSim和真机Z410、Crazyfile产生，主要记录了模型文件数据、飞行日志（虚实结合）数据、飞行日志（外场）数据和飞行日志（3D渲染）数据，数据量共1.88TB。The UAV Swarm Intelligent Behavior Dataset is constructed to support research on the representation of swarm emergent behaviors and machine learning-based recognition technologies, as well as to meet the training and validation requirements of virtual-physical hybrid reinforcement learning. It is generated using the virtual-physical hybrid UAV behavior recognition, training and validation tool UXVSim, as well as real drones Z410 and Crazyfile. The dataset mainly includes four categories of data: model file data, flight logs (hybrid virtual-physical), flight logs (field test), and flight logs (3D rendering), with a total size of 1.88 TB.

搜集汇总

背景与挑战

背景概述

无人机群智行为数据集是一个面向群智涌现行为表示与机器学习识别技术研究的大规模数据集，基于无人机虚实结合行为识别工具和真机产生，包含模型文件和多种飞行日志数据，数据量达1.03TB，支持国家重点研发计划项目“面向群体行为的群智激发汇聚研究”的验证需求。