室内动态环境下的移动机器人自主避障策略

Autonomous obstacle avoidance strategy for mobile robots in indoor dynamic environment

针对室内未知动态环境移动机器人自主避障问题,提出一种融合动态障碍物方向判断策略及子目标点更新策略的自适应模糊神经网络优化避障算法,并依据该算法设计移动机器人避障控制系统。首先,分析移动机器人的运动模型,获取机器人的目标角度;然后由超声波传感器获取障碍物距离信息,由障碍物距离信息判断动态障碍物运动方向并更新子目标点;最后利用自适应模糊神经推理系统实时输出机器人的转向角与速度,实现对机器人转向角的控制,使机器人能够无碰撞地到达目标点。研究结果表明:本文提出的算法能够使移动机器人在未知动态环境下识别障碍物、判断动态障碍物的运动方向以实现自主避障;相对于无子目标点更新策略,移动机器人平均移动速度提高11.75%,验证了所提算法的有效性。

Aiming at the problem of autonomous obstacle avoidance for mobile robots in unknown indoor dynamic environment, an adaptive fuzzy neural network optimization obstacle avoidance algorithm was proposed, which integrates the dynamic obstacle direction judgment strategy and the sub-target point updating strategy. The obstacle avoidance control system of mobile robots was designed based on this algorithm. Firstly, the motion model of mobile robot was analyzed to obtain the target angle of the robot. Then, the distance information of obstacles was obtained by ultrasonic sensor, and the moving direction of dynamic obstacles was judged by the distance information of obstacles. The sub-target points were thus updated. Finally, the steering angle and speed of the robot were displayed in real time by using adaptive fuzzy neural inference system. The robot's steering angle was controlled to reach the target point without collision. The results show that the proposed algorithm enables mobile robots to recognize obstacles and judge the motion direction of dynamic obstacles in an unknown dynamic environment to achieve autonomous obstacle avoidance. Compared with the strategy of no sub-target point updating, the average moving speed of mobile robots is increased by 11.75%, which verifies the effectiveness of the proposed algorithm.