Li Jian, Gao Yuan, Li Zheng, Zhang Weijian, Yu Weilin, Hu Yating, Liu He, Li Changtian
College of Information Technology, Jilin Agricultural University, Changchun, China.
Jilin Province Cross-Regional Collaborative Innovation Center for Agricultural Intelligent Equipment, Jilin Agricultural University, Changchun, China.
Front Plant Sci. 2025 Aug 22;16:1650007. doi: 10.3389/fpls.2025.1650007. eCollection 2025.
Traditional path planning algorithms often face problems such as local optimum traps and low monitoring efficiency in agricultural UAV operations, making it difficult to meet the operational requirements of complex environments in modern precision agriculture. Therefore, there is an urgent need to develop an intelligent path planning algorithm. To address this issue, this study proposes an improved Informed-RRT* path planning algorithm guided by domain-partitioned A* algorithm. The proposed algorithm employs a multi-level decomposition strategy to intelligently divide complex paths into a sequence of key sub-segments, and uses an adaptive node density allocation mechanism to dynamically respond to changes in path complexity. Finally, a dual-layer optimization framework is constructed by combining elliptical heuristic sampling with dynamic weight adjustment. Complex maps are constructed in simulation to evaluate the algorithm's performance under varying obstacle densities. Experimental results show that, compared to traditional RRT* and its improved variants, the proposed algorithm reduces computation time by 56.3%-92.5% and shortens path length by 0.42%-8.5%, while also demonstrating superior path smoothness and feasibility, as well as a more balanced distribution of search nodes. Comprehensive analysis indicates that the A*-MSRRT* (A*-Guided Multi-stage Bidirectional Informed-RRT*) algorithm has strong potential for application in complex agricultural environments.
传统路径规划算法在农业无人机作业中常面临局部最优陷阱和监测效率低等问题，难以满足现代精准农业复杂环境的作业要求。因此，迫切需要开发一种智能路径规划算法。针对这一问题，本研究提出了一种由领域划分A算法引导的改进型有向RRT路径规划算法。该算法采用多级分解策略将复杂路径智能地划分为一系列关键子段，并使用自适应节点密度分配机制动态响应路径复杂度的变化。最后，通过结合椭圆启发式采样和动态权重调整构建了双层优化框架。在仿真中构建复杂地图以评估该算法在不同障碍物密度下的性能。实验结果表明，与传统RRT及其改进变体相比，该算法的计算时间减少了56.3%-92.5%，路径长度缩短了0.42%-8.5%，同时还表现出卓越的路径平滑性和可行性，以及搜索节点分布更加均衡。综合分析表明，A-MSRRT*（A引导的多阶段双向有向RRT）算法在复杂农业环境中具有很强的应用潜力。
