基于多步长蚁群算法的移动机器人路径规划

Research of Robot Path Planning Based on Multi-Step Ant Colony Algorithm

针对蚁群算法在二维栅格环境下的收敛速度慢,规划路径较长等问题,提出一种改进的多步长蚁群算法。首先,通过栅格法对移动机器人的运动环境进行建模;其次,将栅格地图划分为不同的区域,引入人工势场法的思想以及根据栅格间的距离信息对不同区域进行初始信息素浓度优化分布,避免了蚁群盲目搜素,提升了算法的收敛速度;在自适应多步长蚁群算法的基础上,引入了Informed RRT的随机采样区域的思想,缩小了蚁群的搜素空间,提升了路径搜素效率,并同时提升了路径规划的质量;最后,通过MATLAB搭建的模拟仿真环境进行实验。实验结果表明,改进的多步长蚁群算法相比于其他改进算法具有更优的路径和算法效率。

Aiming at the problems of slow convergence speed and long planning path of ant colony algorithm in two-dimensional grid environment,an improved multi-step ant colony algorithm is proposed.Firstly,the moving environment of mobile robot is modeled by grid method.Then,the grid map is divided into different regions,and the idea of artificial potential field method is introduced to optimize the initial pheromone concentration distribution of different regions according to the distance information between grids,which avoids the aimless search of ants and improves the convergence speed of the algorithm.Secondly,based on the adaptive multi-step ant colony algorithm,the idea of random sampling area of Informed RRT is introduced,which reduces the search space of the ant colony,improves the search efficiency of path search,and improves the quality of path planning.Finally,the experiment is carried out through the simulation environment built by MATLAB.The experimental results suggest that the improved multi-step ant colony algorithm has better path quality and algorithm efficiency than other improved algorithms.