DDCNN:一种用于仿真到现实无人机故障诊断的有前景工具
机器人学
2024-06-25 v2
摘要
识别螺旋桨故障对保持四旋翼安全高效运行十分重要。仿真到现实(sim-to-real)无人机故障诊断方法为检测螺旋桨故障提供了一种经济且安全的途径。然而,由于仿真与现实之间的差距,用仿真数据训练的分类器在真实飞行中通常表现不佳。在本工作中,提出一种新颖的基于差值的深度卷积神经网络(DDCNN)模型以解决上述问题。它利用深度卷积神经网络提取的差值特征来缩小仿真到现实的差距。此外,提出一种新的域适应(DA)方法,以进一步使真实飞行数据的分布更接近仿真数据的分布。实验结果表明,DDCNN+DA模型可将真实世界无人机故障检测的准确率从52.9%提升至99.1%。
引用
@article{arxiv.2302.08117,
title = {DDCNN: A Promising Tool for Simulation-To-Reality UAV Fault Diagnosis},
author = {Wei Zhang and Shanze Wang and Junjie Tong and Fang Liao and Yunfeng Zhang and Xiaoyu Shen},
journal= {arXiv preprint arXiv:2302.08117},
year = {2024}
}
备注
20xx IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, including reprinting/republishing this material for advertising or promotional purposes, collecting new collected works for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works