孙国祥副教授团队 | 设施农业自主移动装备导航技术研究进展

摘要：设施农业自主移动装备导航技术的研发对于设施农业的智能化升级具有举足轻重的意义，其面临的核心挑战在于如何在设施场景内部实现安全、高效和精确的移动。通过深入剖析国内外最新的研究动态，本研究旨在全面阐述设施农业自主移动装备导航技术的发展脉络，并探讨其未来的发展趋势。首先梳理了设施农业自主移动装备导航技术的研究现状，重点分析了不同导航方式的优缺点，包括导轨式、循迹式以及基于实时定位与地图构建（simultaneous localization and mapping，SLAM）技术的多源数据融合导航。在此基础上，进一步探讨了基于SLAM技术的多源数据融合导航在实时定位、地图构建以及路径规划等方面的最新研究进展。相关研究表明，传统的导轨式和循迹式导航方式虽然简单易行，但存在路线固定、安装成本高、作业范围受限等问题。相比之下，基于SLAM技术的多源数据融合导航则具有更高的灵活性和自主性，能够实现精准定位、智能避障以及自适应路径规划等功能，因此在设施农业自主移动装备领域具有广阔的应用前景。随着SLAM技术的不断发展和完善，基于多源数据融合导航的设施农业自主移动装备将迎来更加广阔的发展空间。同时，本研究也关注到这些技术在实际应用中可能存在的挑战和限制，如数据处理能力、算法优化等问题。最后，对我国设施农业自主移动装备导航技术的研发提出了几点展望：采用多源数据融合式导航技术，实现高精度定位与地图构建；引入人工智能、云计算与边缘计算等技术，提升自主移动装备对于复杂设施场景的识别和理解能力；通过全局路径规划和局部路径规划相结合，实现自主移动装备的自主决策和路径规划；通过智能避障技术，提高自主识别和避让障碍物能力，确保作业过程中的安全性与可靠性；深入研究多机协同、人机协同的自主移动装备，实现多任务协同作业，提升装备适用性和作业效率。该研究旨在为我国设施农业自主移动装备智能化发展提供有益的参考和借鉴。

关键词：设施农业, 导航, 自主移动装备, 实时定位, 地图构建, 路径规划

Abstract: The development of navigation technologies for autonomous mobile equipment is pivotal to the intelligent upgrading of facility agriculture， with its core challenge being how to achieve safe， efficient， and precise movement within greenhouse environments. Through an in-depth analysis of the latest research trends， this study aims to comprehensively elucidate the evolution of these navigation technologies and discuss their future directions. We begin by reviewing the state of the art， with a focus on comparing the advantages and disadvantages of different navigation methods， including guide-rail， path-tracking， and multi-source data fusion navigation based on simultaneous localization and mapping （SLAM） technology. Furthermore， we explore recent advances in SLAM-based multi-source data fusion navigation for real-time localization， mapping， and path planning. Relevant research indicates that while traditional guide-rail and path-tracking methods are simple to implement， they suffer from fixed routes， high installation costs， and limited operational range. In contrast， SLAM-based multi-source data fusion navigation offers greater flexibility and autonomy， enabling precise localization， intelligent obstacle avoidance， and adaptive path planning， thus holding broad application prospects in facility agriculture. As SLAM technology continues to mature， equipment employing this navigation approach will see even wider adoption. This study also addresses practical challenges and limitations， such as computational demands and algorithm optimization. Finally， we propose several development prospects for China's facility agriculture： adopting multi-source data fusion navigation to achieve high-precision localization and mapping； integrating artificial intelligence， cloud computing， and edge computing to enhance the equipment's perception and understanding of complex environments； combining global and local path planning to enable autonomous decision-making； implementing intelligent obstacle avoidance to ensure operational safety and reliability； and investigating multi-vehicle and human-vehicle collaboration for efficient multi-task coordination. This research aims to provide valuable insights and references for the intelligent development of autonomous mobile equipment in Chinese facility agriculture.

Keywords: facility agriculture, navigation, autonomous walking equipment, real-time positioning, map construction, path planning