激光数据聚类和Morphin算法下的机器人避障研究

Research on Obstacle Avoidance of Mobile Robot Based on Laser Data Clustering and Morphin Algorithm

为了解决移动机器人在未知环境下的避障问题,提出了一种从障碍物检测、预测到避撞的避障方法。设定圆形窗口作为机器人有效扫描区域,利用激光传感器采集到的数据结合聚类、匹配和分类算法确定障碍物类型和动态障碍物运动信息,绘制窗口内的动态局部地图来预测动态障碍物与机器人的碰撞关系,结合Morphin算法实现有效的避障。仿真实验表明,在该算法下移动机器人能够有效地检测出障碍物,进行碰撞预测,并做出合理地避障措施。

In order to solve the problem of obstacle avoidance for mobile robot in unknown environment,a method of obstacle avoidance including obstacle detection and prediction is proposed.Setting the circular window as an effective scanning area of robot,combing with clustering,matching and classification algorithm to determine the motion types of obstacles and motion information of dynamic obstacles by laser sensor information collected data,draw a map of local dynamic collision window to predict the collision relation between dynamic obstacles and the robot,and using the Morphin algorithm to achieve obstacle avoidance effectively.The simulation results show that the mobile robot can detect obstacles effectively,predict the collision and make reasonable obstacle avoidance measures.