基于改进蚁群算法的自动导引运输车全局路径规划方法研究

Global path planning for automated guided vehicle based on improved ant colony algorithm

针对自动导引运输车(AGV)全局路径规划采用传统蚁群算法存在收敛速度慢、易陷入局部最优的问题,提出了基于改进蚁群算法的AGV全局路径规划方法。首先,运用MAKLINK图论法构建了具有障碍物的环境模型,作为路径规划的基础;其次,改进的蚁群算法中融合了动态权重目标导向原理,设计了一种新的启发式函数,提高了其选择距离目标点更近的可选节点的概率,减小了AGV对非最短路径的选择概率;然后,采用动态调整信息素挥发系数策略进行了信息素更新,提高了算法的搜索效率;最后,将改进蚁群算法与传统蚁群算法进行了仿真实验对比。研究结果表明:与传统蚁群算法相比较,改进措施可使收敛速度提升近一倍,路径规划效率显著提高。

Aiming at the problems of slow convergence speed and trapping into local minimum of global path planning for automated guided vehicle by traditional ant colony algorithm,a global path planning for AGV based on improved ant colony algorithm was proposed.At first,the environment models with obstacle were established as the basis for path planning by MAKLINK graph.Secondly,the improved ant colony algorithm was combined with dynamic weight goal-oriented principle,then a new heuristic function was designed,to improve the probability of selecting the closer path to the target point,and reduce the probability of selecting the non-shortest path.The pheromone was updated with the strategy of dynamic adjustment of pheromone decay parameter for improving the search efficiency.Finally,the improved ant colony algorithm was compared with the traditional ant colony algorithm by simulation experiment.The results indicate that compared to traditional ant colony algorithm,improvements can increase the convergence speed by nearly one time and improve path planning efficiency.