农业机器人底盘运动控制技术现状与发展趋势

Current Status and Development Trends of Agricultural Robot Chassis Motion Control Technology

摘要:

农业机器人底盘是实现农业装备智能化作业的基础载体，其运动控制能力直接影响机器人在非结构化环境下作业稳定性与可靠性。本文从底盘机械构型、核心控制技术以及典型场景适配策略3个方面，对国内外农业机器人底盘运动控制研究现状进行了系统梳理。分析了轮式、履带式及腿足式底盘的结构特点与运动机理，对比了不同构型在作业效率与地形通过性方面的优势与局限。重点阐述了感知技术、路径规划、驱动控制、路径跟踪及姿态控制等关键技术研究进展，指出当前控制算法正逐步从基于理想几何模型运动学跟随，向适应复杂环境扰动的动力学控制转变。同时，针对大田高速作业、设施狭窄空间及丘陵大坡度3种典型工况，归纳了不同环境约束下差异化控制策略与解决方案。最后，提出了现有技术在模型失配、算力限制及通用性等方面的不足，并对数字孪生仿真、边缘计算轻量化及多源感知融合等方向进行了展望，可为农业机器人相关研究提供参考。

Abstract:

The agricultural robot chassis serves as the fundamental platform for achieving intelligent agricultural operations.Its motion control capability is critical to the operational stability and reliability of robots in unstructured environments.It systematically reviewed the research status of motion control for agricultural robot chassis from three dimensions: mechanical configurations, core control technologies, and adaptation strategies for typical scenarios.Firstly, the structural characteristics and motion mechanisms of wheeled, tracked, and legged chassis were analyzed, comparing the advantages and limitations of different configurations in terms of operational efficiency and terrain trafficability.Secondly, the research progress of key technologies, including perception technology, path planning, drive control, path tracking, and attitude control was elaborated.It was pointed out that current control algorithms were gradually shifting from kinematic following based on ideal geometric models to dynamic control capable of adapting to complex environmental disturbances.Simultaneously, aiming at three typical working conditions, i.e., high-speed field operations, narrow spaces in facilities, and large slopes in hilly areas, the differentiated control strategies and solutions under different environmental constraints were summarized.Finally, the deficiencies of existing technologies in terms of model mismatch, computing power limitations, and versatility were summarized.Future development trends were prospected from the directions of digital twin simulation, lightweight edge computing, and multi-source perception fusion, aiming to provide a reference for related research.