基于改进双层蚁群进化算法的无人机路径规划

UAV Path Planning Based on Improved Double Layer Ant Colony Evolutionary Algorithm

中文关键词:  无人机  路径规划  双层蚁群算法  A^*算法  进化算法

英文关键词:UAV  path planning  double layer ant colony algorithm  A* algorithm  evolutionary algorithms

中文摘要:

      针对蚁群算法在无人机路径规划中存在的算法初期盲目搜索、收敛速度慢、转弯次数过多等问题，提出一种改进的双层蚁群进化算法，将蚂蚁分成引导层和优化层。算法初期搜索阶段使用A^*算法得到次优路径，以该路径非均匀化分配信息素。接着，设计了一个双层搜索的模式，分别设计了双层蚂蚁的启发式函数，基础层寻路，寻优层用于进一步寻找优化路径；信息素更新的过程中引入了进化算法的思想，提高了蚂蚁种群的多样性。信息素更新方式加入了信息素增强削弱因子，从而增强了算法的正反馈机制。最后对规划出的初始路径进行优化，双向删除冗余节点。实验结果表明，相比于传统蚁群算法，所提算法在迭代次数、优化路径长度和转弯次数方面有明显改善。

英文摘要:

      Aiming at the shortcomings of traditional ant colony algorithm in UAV path planning,such as initial blind search,slow convergence speed,and excessive number of turns,an improved double layer ant colony evolutionary algorithm is proposed,which divides ants into a guidance layer and an optimization layer.In the initial search stage of the algorithm,the suboptimal path is obtained by using A^* algorithm,and the pheromones are distributed unevenly along this path.Next,a double layer search pattern is designed.Heuristic functions for the double layers of ants are designed separately.The base layer searches for the arrival path,while the optimization layer further searches for the optimization path.In the process of updating pheromones,the idea of evolutionary algorithms is incorporated to improve the diversity of ant populations.A pheromone enhancement and weakening factor is added to enhance the positive feedback mechanism of ant colony algorithms.Finally,the initial path is optimized to remove redundant nodes in both directions.The experiment results show that compared to the traditional ant colony algorithm,the improved algorithm has significant improvements in terms of iteration times,optimized path length,and turning times.