基于改进 ND＋算法的机器人避障方法设计

摘要

When it comes to the issue of obstacIe avoidance for mobiIe service robots in structured indoor environments,traditionaI Nearness-Diagram+ Navigation (ND+) method cannot aIways output smooth veIocity.Therefore,a new method is proposed to revise the desired direction output by ND+ method.MobiIe robots can thus avoid obstacIes and reach the goaI faster and smoother than before.At first,the map is divided into two parts based on the initiaI desired direction output by ND+ method. Then a point set of obstacIes around the mobiIe robot is generated to revise the desired direction of the robot.%针对室内结构化环境下服务机器人避障过程中传统 ND＋算法生成的决策速度不平滑问题，提出了一种对 ND＋算法生成的决策方向进行修正，使得机器人决策速度变得平滑的设计方法，该方法能够使得机器人更快速平稳地避开障碍物到达目标点。首先，根据 ND＋算法获取机器人的初始决策方向，以此方向把地图划分为两部分，获取左右两侧的近距离的障碍物点集，各个障碍物点共同对机器人作用，修正机器人的决策方向；实验结果表明，相比 ND＋算法，该方法能够提前规避前进方向潜在障碍物，机器人更加平缓地行走于充斥拥挤障碍物的室内环境中到达指定目标点。