Authors

Abstract

A multi-unmanned aerial vehicle (UAV) task allocation and path planning model with the maximum endurance constraint was constructed specific to the agricultura l patrol scene. Moreover, an optimized ant colony optimization (ACO) algorithm applicable to grid map environment was proposed given such problems of the traditional ACO algorithm as limited path search direction and field of view, failure to find the shortest path and proneness to deadlock. This method preprocessed the grid map environment, extracted the feature points of obstacles, and selected such feature points as the way-finding access nodes; then, the construction efficiency of the solution was enhanced via the nonuniform pheromone distribution based on ACO algorithm, the guiding function of path search was strengthened using Tent chaotic mapping, and the pheromone evaporation coefficient was dynamically adjusted to prevent the algorithm from too early convergence. The experimental results show that the proposed method more conforms to the operational requirements of rotary-wing UAVs with limited cruising ability in comparison with the existing methods. Besides, the convergence efficiency of the improved ACO algorithm embedded with the niche genetic algorithm is 30.55% higher than that of the traditional ACO algorithm. The experimental results verify the practicability and effectiveness of the proposed method.

Abstract in Chinese

针对农业巡检场景的多无人机任务分配与路径规划问题，构建一种最大航程约束的多无人机任务分配与路径规划模型。针对传统蚁群算法存在的路径搜索方向和视野受限、无法找到最短路径、容易发生死锁等问题，提出了一种适用于网格地图环境下的优化蚁群算法。该方法对网格地图环境进行预处理，提取障碍物的特征点，并选择这些特征点作为寻路访问节点; 然后，基于蚁群算法，采用信息素不均匀分布来提高解的构造效率，采用Tent混沌映射增强路径搜索的引导作用，动态调整信息素挥发系数以避免算法过早收敛。实验结果表明，提出的方法相比于现有方法更符合续航能力有限的旋翼无人机作业需求，且相比于传统蚁群算法，提出的嵌入小生境遗传算法的改进蚁群算法与传统蚁群算法相比，算法收敛效率提升30.55%。实验结果证明了所提方法的实用性和有效性。