激光雷达和神经网络相融合的机器人避障轨迹规划研究  

Obstacle avoidance trajectory planning of robot based on combination of lidar and neural network

作　　者：刘祎[1] 刘萍[1] 李守军[1] LIU Yi;LIU Ping;LI Shoujun(Suqian College Institute of Electrical and Mechanical Services,Suqian Jiangsu 223800,China)

机构地区：[1]宿迁学院机电工程学院,江苏宿迁223800

出　　处：《激光杂志》2021年第9期175-178,共4页Laser Journal

基　　金：江苏省高等学校自然科学研究项目(No.19KJB440002);江苏省宿迁市科技局指导性项目(No.Z2018229);江苏省宿迁市科技局指导性项目(No.Z2019108)。

摘　　要：机器人避障轨迹规划研究当前是人工智能研究领域的热点,当前机器人避障轨迹规划方法对环境信息敏感,机器人避障轨迹规划偏差大,在短时间难以找到最优机器人避障轨迹规划结果,为了获得理想的机器人避障轨迹规划结果,提出了基于激光雷达和神经网络相融合的机器人避障轨迹规划方法。首先分析当前机器人避障轨迹规划方法的局限性,并采用测量范围广、精度高的激光雷达对机器移动的环境信息,然后将环境信息作为BP神经网络的输入,对环境信中障碍物进行分类,最后根据障碍物分类结果采用人工势场方法找到机器人避障轨迹规划结果,并与其他机器人避障轨迹规划方法进行了对比测试,结果表明,本方法可以在短时间内找到最优的机器人移动最优避障轨迹规划路径,可以避开各种障碍物,保障机器人安全到达目的地,相对对比方法,本方法获得更理想的机器人移动最优避障轨迹规划结果。At present,robot obstacle avoidance trajectory planning is a hot topic in the field of artificial intelligence. The current robot obstacle avoidance trajectory planning method is sensitive to environmental information,and the robot obstacle avoidance trajectory planning deviation is large,so it is difficult to find the optimal robot obstacle avoidance trajectory planning results in a short time. In order to obtain the ideal robot obstacle avoidance trajectory planning results,this paper proposes a method based on lidar and neural network. The method of obstacle avoidance trajectory planning is proposed. Firstly,the limitations of the current robot obstacle avoidance trajectory planning methods are analyzed,and the environment information of the machine movement is analyzed by using the lidar with wide measurement range and high precision. Then,the environment information is used as the input of BP neural network to classify the obstacles in the environmental information. Finally,according to the classification results of the obstacles,the artificial potential field method is used to find the robot obstacle avoidance trajectory planning results. The results show that this method can find the optimal path of Obstacle Avoidance Trajectory Planning for robot movement in a short time,avoid all kinds of obstacles and ensure the robot to reach the destination safely. Compared with other methods,this method obtains more ideal optimal obstacle avoidance trajectory planning results.

关 键 词：未知环境 机器人避障 激光雷达 轨迹规划 验证性测试 环境信息 

分 类 号：TN929[电子电信—通信与信息系统]