Abstract: In order to overcome the influence of traditional ant colony algorithm which is easy to fall into local optimum and slow convergence speed, a simulation model of robot experimental environment is established by using grid map. The ant colony algorithm is improved and applied to the robot path planning. Firstly, considering the directionality from the starting point of the path planning to the target point, the easy to fall into the local optimal solution in the early stage and the convergence speed of the ant, a heuristic function adding the bidirectional search direction mechanism and the proportional coefficient guiding factor is proposed to avoid the algorithm selecting the area that runs away from the end direction or the disadvantage of walking the loop. Secondly, according to the different times of different road segments being selected, different pheromone weights are set, the importance of different road segments is strengthened, and the convergence speed of the algorithm is accelerated. Finally, the algorithm simulation is carried out on the matlab software platform. The simulation results verify the effectiveness of the proposed method.

Key words: ant colony algorithm, heuristic function, pheromone weight, convergence speed

摘要： 为了克服传统蚁群算法易陷入局部最优且收敛速度慢的影响，采用栅格地图建立机器人实验环境仿真模型。针对蚁群算法进行改进并将其应用到机器人路径规划上。考虑到从路径规划起点到目标点的方向性、前期存在的易陷入局部最优解以及蚂蚁收敛速度的问题，提出了添加双向搜索方向机制和比例系数引导因子的启发函数，避免了算法在搜索过程中选择与终点方向相背的区域行走或者走回路的弊端。根据不同路段被选择次数不同，设置不同信息素权重，强化了不同路段的重要性，加快算法收敛速度。在matlab软件平台上进行算法仿真，仿真结果验证了该方法的有效性。

关键词: 蚁群算法, 启发函数, 信息素权重, 收敛速度