基于视觉识别和路径规划的搬运机器人设计  

Design of Handling Robot Based on Visual Recognition and Path Planning

作　　者：卜令宇 魏领会[1] 熊壮 白晨 张大正 袁铮铮 王文韬 

机构地区：[1]北方工业大学,北京

出　　处：《计算机科学与应用》2023年第2期204-218,共15页Computer Science and Application

摘　　要：本文设计了一台基于视觉识别和自动路径规划的搬运机器人,该机器人以树莓派4B为主控,以ROS为开发平台。通过单目相机和激光测距模块来识别物料的颜色和空间位置,推导出了逆运动学模型来控制机械臂抓取物块。机器人移动过程中采用激光测距模块和电子罗盘进行全场定位和姿态矫正,并自动规划运动路径。PID算法保证驱动电机不受电源电压影响。经过测试验证,该机器人能够快速地规划出合理的路径,准确识别物体并进行抓取和投送。This paper designs a handling robot based on visual recognition and automatic path planning. The robot is controlled by Raspberry Pi 4B and ROS is the development platform. The color and spatial position of the material are identified through the monocular camera and the laser ranging module, and an inverse kinematics model is derived to control the robotic arm to grab the object. During the movement of the robot, the laser ranging module and the electronic compass are used to perform full-field positioning and attitude correction, and automatically plan the movement path. The PID algorithm ensures that the drive motor is not affected by the power supply voltage. After testing and verification, the robot can quickly plan a reasonable path, accurately identify objects, and grasp and deliver them.

关 键 词：搬运机器人 运动学 颜色识别 视觉定位 路径规划 

分 类 号：TP2[自动化与计算机技术—检测技术与自动化装置]