多无人机协同实现精准农业全流程自动化
Architecture for Multi-Unmanned Aerial Vehicles based Autonomous Precision Agriculture Systems
- 构建多无人机协作的农业任务架构，支持自主规划与执行
- 实测显示系统具备容错性、鲁棒性及高效数据处理能力
- 适合需要无人化农田管理的农场与农业科技开发者
近年来，无人机在精准农业中的应用迅速增长。为支持田间各类算法的部署，本文定义了无人机在精准农业中的各项任务，并将其建模为一套抽象架构。该架构基于无需频繁人工干预的前提，通过多架协同作业的无人机完成图像处理、路径规划、通信、数据采集和田间地图生成等任务，实现高效运行。设计中充分考虑了多无人机在农业应用中的各类限制因素。系统提供从任务规划、数据采集到图像处理的全流程自主解决方案，经仿真与田间测试验证，具备故障容错、鲁棒性强、开发与使用友好等优势。
原文摘要 · Abstract (English)
The use of unmanned aerial vehicles (UAVs) in precision agriculture has seen a huge increase recently. As such, systems that aim to apply various algorithms on the field need a structured framework of abstractions. This paper defines the various tasks of the UAVs in precision agriculture and model them into an architectural framework. The presented architecture is built on the context that there will be minimal physical intervention to do the tasks defined with multiple coordinated and cooperative UAVs. Various tasks such as image processing, path planning, communication, data acquisition, and field mapping are employed in the architecture to provide an efficient system. Besides, different limitation for applying Multi-UAVs in precision agriculture has been considered in designing the architecture. The architecture provides an autonomous end-to-end solution, starting from mission planning, data acquisition and image processing framework that is highly efficient and can enable farmers to comprehensively deploy UAVs onto their lands. Simulation and field tests shows that the architecture offers a number of advantages that include fault-tolerance, robustness, developer and user-friendliness.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。