Abstract: In view of the deficiencies existed in global path planning algorithms such as goal nonreachable problem easily occurred, low computational efficiency and poor smoothness of the planned path, an optimal global path planning algorithm based on new intelligent optimization algorithm - firefly algorithm combined with traditional artificial potential field method is proposed. Firstly, according to the known global map, the artificial potential field method with simple algorithm structure, smaller amount of calculation and faster operation speed is used as the initialization guidance factor to initialize the firefly algorithm parameters. Then, in the polar coordinate system replacing the rectangular coordinate system, the firefly algorithm is used to get the optimization solution of the planning path to realize the global path planning. The experimental results show high efficiency, good effect in obstacle avoidance, and high safety and reliability of the proposed algorithm, and the optimal path is closer to the ideal path, the average error is below 0.08 m, which can accomplish the goal of path planning well.

Key words: global path planning for robot, artificial potential field method, improved firefly algorithm, polar coordinate system

摘要： 鉴于已有机器人全局路径规划算法存在的易出现目标不可达问题、计算效率较低、所规划路径平滑性欠佳等不足，提出将新型智能优化算法——萤火虫算法与传统人工势场法相结合的最优全局路径规划算法。首先，根据已知全局地图，利用算法结构简单、计算量较小、运算速度较快的人工势场法作为初始化引导因子对萤火虫算法参数进行初始化，再以极坐标系代替直角坐标系利用萤火虫算法对规划路径进行寻优求解。极坐标法能自动舍弃规划路径上的冗余点，增强所规划路径的平滑性；同时考虑到传统萤火虫算法所存在的不足，对萤火虫算法进行了相应的改进：引入自适应步长改进随机步长，加快算法的收敛，并以混沌逻辑改进萤火虫算法的吸收系数，避免算法收敛到局部最优，在一定程度上解决人工势场法所存在局部震荡和目标不可达问题。实验结果表明所提算法实现效率高、避障效果好，安全可靠性好，最优路径更加接近理想路径，平均误差在0.08 m以下，能够很好地完成路径规划的目标。

关键词: 全局路径规划, 人工势场法, 改进萤火虫算法, 极坐标系