高级检索
    王伟, 张彦斐, 宫金良. 基于蚁群−BFS算法的复杂环境下农业机器人全区域覆盖研究[J]. 华南农业大学学报, 2021, 42(3): 119-125. DOI: 10.7671/j.issn.1001-411X.202009027
    引用本文: 王伟, 张彦斐, 宫金良. 基于蚁群−BFS算法的复杂环境下农业机器人全区域覆盖研究[J]. 华南农业大学学报, 2021, 42(3): 119-125. DOI: 10.7671/j.issn.1001-411X.202009027
    shu
    WANG Wei, ZHANG Yanfei, GONG Jinliang. Study on the whole area coverage of agricultural robot in complex environment based on ant colony-BFS algorithm [J]. Journal of South China Agricultural University, 2021, 42(3): 119-125. DOI: 10.7671/j.issn.1001-411X.202009027
    Citation: WANG Wei, ZHANG Yanfei, GONG Jinliang. Study on the whole area coverage of agricultural robot in complex environment based on ant colony-BFS algorithm [J]. Journal of South China Agricultural University, 2021, 42(3): 119-125. DOI: 10.7671/j.issn.1001-411X.202009027
    shu

    基于蚁群−BFS算法的复杂环境下农业机器人全区域覆盖研究

    Study on the whole area coverage of agricultural robot in complex environment based on ant colony-BFS algorithm

      摘要
    • 摘要:
      目的  以路径重复率为优化目标解决农业机器人在数字生态农场中的全区域覆盖问题。
      方法  首先，将栅格地图中的障碍物进行膨胀处理，在此基础上进行矩形分区以及分区合并操作；然后，通过改进的蚁群算法规划分区间的遍历顺序、通过改进的广度优先搜索(Breadth first search, BFS)算法规划分区间终点与起点的衔接路径，从而实现机器人全区域覆盖。2种算法的具体改进方案为：分别通过人工免疫算法与粒子群算法改进遗传算法的选择与交叉算子，并将改进后的选择算子、交叉算子、原遗传算法变异算子与蚁群算法相结合改进传统蚁群算法信息素更新方法；建立动态函数以简化BFS算法规划的路径。
      结果  仿真结果表明，改进蚁群算法收敛时的迭代次数较传统蚁群算法减少了83.1%，路径长度相比减少了4.8%；由改进的蚁群算法与改进的BFS算法规划的机器人遍历路径重复率是传统蚁群算法和BFS算法的56%，且农业机器人能实现对农田区域的100%覆盖。
      结论  本研究提供了一种农业机器人在复杂环境的数字生态循环农场中进行全遍历覆盖的解决方案。

       

      Abstract:
      Objective  To solve the whole area coverage problem of agricultural robot in digital ecological farm with path repetition rate as the optimization target.
      Method  Firstly, the obstacles in the raster map were expanded and rectangular partitioning was performed and the partitions were merged. Then, the improved ant colony algorithm was used to calculate the traversal order among partitions, and the improved breadth first search (BFS) algorithm was used to calculate the connecting path between the end point and the starting point between partitions. By this way, the whole area coverage of robot could be achieved. The improvement schemes of the two algorithms were as following: The selection and crossover operators of genetic algorithm were improved by artificial immune algorithm and particle swarm optimization algorithm respectively. The improved selection and crossover operators, mutation operator of original genetic algorithm and ant colony algorithm were combined to improve the pheromone updating method of traditional ant colony algorithm. The dynamic function was established to simplify the path planned by BFS algorithm.
      Result  The simulation results showed that the number of iterations and path length of the improved ant colony algorithm were 83.1% and 4.8% less than those of the traditional ant colony algorithm. The path repetition rate of the improved ant colony algorithm and the improved BFS algorithm was 56% of the traditional ant colony algorithm and BFS algorithm, and the agricultural robot could achieve 100% coverage of farmland.
      Conclusion  This study provides a solution for agricultural robots to perform the whole area coverage in digital ecological circular farm with complex environments.

       

      • HTML全文
      • 参考文献(19)
      • 相关文章
      • 施引文献
    • 资源附件(0)

    /

    返回文章
    返回