Authors

Abstract

To address the challenge of balancing obstacle avoidance and flight safety in 3D route planning for fixed-wing UAVs performing crop irrigation in mountainous agricultural areas, this study proposed an improved multi-strategy dung beetle optimization (MDBO) algorithm. Specifically, constraints on the safe turning radius and pitch angle of fixed-wing UAVs were set to ensure stable flight of UAVs in the complex terrain of agricultural and forestry mountainous areas (including crop plots, tree obstacles, gullies, etc.), avoid collisions, and guarantee the continuity of irrigation operations. Meanwhile, the golden sine strategy, Bernoulli chaotic mapping, dynamic adaptive weight strategy, and adaptive Gaussian-Cauchy mixed perturbation mutation strategy were introduced to enhance the algorithm’s local exploration capability, global search ability, and convergence performance in irrigation path optimization for agricultural and forestry mountainous areas. By comparing the performance of the improved MDBO with the basic dung beetle optimization (DBO) algorithm using 9 benchmark test functions, the results show that MDBO has a faster convergence speed and stronger stability. Through simulation experiments on irrigation route planning for crops in agricultural and forestry mountainous areas, it is verified that MDBO outperforms traditional algorithms in irrigation path accuracy (coverage uniformity, path length optimization) and convergence performance, and can meet the route requirements for efficient irrigation of crops in agricultural and forestry mountainous areas.

Abstract in Chinese

为解决山区农业区固定翼无人机进行作物灌溉时三维路径规划中避障与飞行安全的平衡难题，本研究提出了一种改进的多策略蜣螂优化（MDBO）算法。具体而言，设定了固定翼无人机安全转弯半径和俯仰角的约束条件，以确保无人机在农业和林业山区复杂地形（包括农田、树木障碍物、沟壑等）中稳定飞行，避免碰撞，并保证灌溉作业的连续性。同时，引入了黄金正弦策略、伯努利混沌映射、动态自适应权重策略和自适应高斯-柯西混合扰动变异策略，以增强算法在农业和林业山区灌溉路径优化中的局部探索能力、全局搜索能力和收敛性能。通过使用 9 个基准测试函数将改进的 MDBO 算法与基本的蜣螂优化（DBO）算法的性能进行比较，结果表明 MDBO 具有更快的收敛速度和更强的稳定性。通过对农林山区农作物灌溉路线规划的模拟实验，验证了 MDBO 在灌溉路径精度（覆盖均匀度、路径长度优化）和收敛性能方面优于传统算法，能够满足农林山区农作物高效灌溉的路线要求。