农用无人机自主飞行技术研究与趋势

Research and trend of autonomous flight technology of agricultural UAV

随着人工智能、深度学习的快速发展,无人机自主飞行技术已然成为无人机智能化评价之一。在精准农业与智慧农业的倡导下,无人机在农业领域发展迅猛。农用无人机应用场景包括:作物授粉、喷洒作业、农情监测、打顶剪枝、疏花疏果及畜牧跟踪,其自主程度重要性不言而喻。综述国内外无人机飞行技术研究现状,介绍农用无人机在自主控制方法、避障方法、轨迹规划算法以及精准喷洒方法的研究进展,分析指出农用无人机系统自主环境感知能力差、信息处理速度慢、路径规划算法收敛慢、作物识别率低等不足,提出采用多传感器组合、双冗余控制、多算法融合和基于深度学习的作物特征识别等改进方法。本文为农用无人机自主飞行技术满足智能作业需求提供理论基础。

With the rapid development of artificial intelligence and deep learning,autonomous flight technology of UAV has become one of the intelligent evaluation of UAV.Under the guidance of precision agriculture and intelligent agriculture,UAV is developing rapidly in the field of agriculture.The application scenarios of agricultural UAV include:crop pollination,spraying operation,agricultural situation monitoring,toppling and pruning,flower and fruit thinning and livestock tracking,and the importance of their autonomy is self-evident.It was summarized the research status of UAV flight technology at home and abroad,and introduced the research progress of agricultural UAV:method of autonomous control,method of obstacle avoidance,trajectory planning algorithm and the research progress of precision spraying method.It was pointed out that agricultural UAV system autonomous environmental awareness was poor,such as poor sense of autonomous environment,slow speed of information processing,slow convergence of path planning algorithm,low rate of crop recognition.The improved methods such as multi-sensor combination,dual-redundancy control,multi-algorithm fusion and crop feature recognition based on depth learning were proposed.It provides a theoretical basis for autonomous flight technology of Agricultural UAV to meet the requirements of intelligent operation.