基于CNN的移动机器人局部路径激光雷达辅助规划方法

LiDAR-assisted planning method for local path of mobile robot based on CNN

移动机器人在开展局部路径规划时,对障碍物的准确识别及避障路径规划效果会直接影响机器人的安全运行。为此,提出基于CNN的移动机器人局部路径激光雷达辅助规划方法。该方法首先依据CNN方法建立障碍物定位模型,结合设计的视觉控制器,完成场地内障碍物位置的定位;再使用激光雷达采集机器人与障碍物位置之间的距离,再通过VFH算法对距离量化,利用自适应阈值计算机器人移动时的转向,从而确定机器人的局部航向,实现机器人的局部路径规划。实验结果表明,使用该方法实施路径规划时,规划出的路径长度最多为82 m,规划时间最多为5.12 s,规划效果好。

When a mobile robot carries out local path planning,the accurate identification of obstacles and the effect of obstacle avoidance path planning will directly affect the safe operation of the robot.Therefore,a local path lidar-assisted planning method of mobile robot based on CNN is proposed.Firstly,the obstacle positioning model is established according to the CNN method,and the location of obstacles in the site is completed by combining the designed visual controller.Then,lidar is used to collect the distance between the robot and the obstacle position,and then the distance is quantified by the VFH algorithm,and adaptive threshold is used to calculate the turning direction of the robot when it moves,so as to determine the local course of the robot and realize the local path planning of the robot.The experimental results show that when using this method to implement path planning,the planned path length is 82 m at most,and the planning time is 5.12 s at most,and the planning effect is good.