基于改进蚁群算法的智能小车路径规划方法

A New Path Planning Method of Intelligent Car Based on Ant Colony Algorithm

针对基于蚁群算法的智能小车在复杂环境下易陷入局部最优的路径规划问题,提出一种改进蚁群算法的全局动态路径规划方法。该方法依据智能小车行驶的环境信息,首先通过栅格法进行环境建模,以初始信息素的不均匀分配方法代替信息素的均匀分配方式,避免了经典算法中蚂蚁走回路的问题;其次调整启发函数,为蚂蚁后续的路径选择提供信息;然后改进蚁群算法中局部信息素和全局信息素的更新规则,进一步加快了路径规划算法的收敛速度;最后在转移概率中增加邻域安全因子,提高路径规划算法的效率;最终以智能小车为对象,在设置不同复杂度和障碍物规模的情况下,对智能小车进行全局动态路径规划的仿真实验,实验结果证明本方法具有实时性与高效性。

Aiming at the path planning of intelligent car based on ant colony algorithm which is easy to fall into the local optimization in the complex environment,a global dynamic path planning method is proposed.According to the environment information,the environment map is built by raster firstly.Secondly,the uneven distribution method of the initial pheromone is used to replace the uniform distribution of the pheromone,which avoids the problem of the ant circuit.Thirdly,the local pheromone and global pheromone in ant colony algorithm are optimized to accelerate the convergence speed of path planning algorithm.Fourthly,the heuristic factor information on the path is updated to optimize the subsequent path selection of ants.Finally,taking the intelligent car as object,the proposed algorithm is verified under the environmental condition of different complexity.