一种基于模糊识别的移动机器人避障算法  

Obstacle avoidance algorithm for mobile robot based on fuzzy identification

作　　者：吴康[1] 沈捷[2] 费树岷[1] 

机构地区：[1]东南大学自动化学院,南京210096 [2]南京工业大学自动化学院,南京210009

出　　处：《东南大学学报（自然科学版）》2008年第A02期199-202,共4页Journal of Southeast University：Natural Science Edition

基　　金：国家自然科学基金资助项目(60574006)

摘　　要：设计了一种基于障碍物模糊识别的移动机器人避障方案.移动机器人的视觉导航,具有信息量大的特点,在避障中相对其他传感系统,有一定优势.视觉图像处理计算量大复杂度高,避障中采用模糊控制的思想,减少图像矩阵精确变化和处理,选取障碍物图像中部分边缘点作为特征坐标点,据此设计模糊控制算法识别出障碍物其大致的姿态信息.根据障碍物图像大致姿态信息和其在视觉探测图像区域的方位信息,设定一组虚拟目标点序列,合理规划移动机器人的安全避障路径.最后,仿真试验演示方案良好的避障效果,证明在简单视觉环境中可行性和有效性.An obstacle avoidance scheme of mobile robot based on identification of obstacle has been designed according to the theory of fuzzy control. It is based on vision information therefore is more effective and may have more information of obstacle than other sensor system. Since the process is more complicated, the general stance of the obstacle is obtained through some special points of the obstacle image side by using fuzzy identification. The rational layout of the obstacle avoidance route is accomplished base on the location and general stance of the obstacle, using the method of enacting a set of virtual target point. The feasibility and effectiveness of this scheme have been proved by simulation study and experiment.

关 键 词：移动机器人 避障 模糊识别 虚拟目标 

分 类 号：U459.2[建筑科学—桥梁与隧道工程]