基于ROS和EtherCAT的视觉伺服7DOF机械臂仿真研究

Research on vision servo 7DOF manipulator simulation based on ROS and EtherCAT

将视觉传感器构建到机械臂一侧的视觉伺服系统是7自由度机械臂实现物体抓取和移动的有效方法,但此方法存在着对目标识别定位速度慢和末端执行器轨迹规划难的问题。因此,设计并实现了基于机器人操作系统(Robot Operating System,ROS)和EtherCAT的视觉伺服7自由度机械臂仿真平台。在ROS系统上,利用颜色机器视觉(Color Machine Vision,CMVision)算法实现了抓取目标的快速识别和定位;基于机械臂的D-H模型,求解得到逆运动方程,规划机械臂的轨迹,并应用比例、积分、微分(Proportion,Integration,Differentiation,PID)算法进行控制;在Linux上构建实时内核,利用EtherCAT技术实现了机械臂末端执行器的快速准确移动控制。测试结果表明,该平台能够实现目标的快速识别、定位和抓取,具备良好的稳定性和可操作性。

The visual servo system integrating the visual sensor into manipulator side of the mechanical arm is an effective method to achieve grasping and moving of the object by 7 Degrees of Freedom（ 7 DOF）manipulator. However,the low speed during positioning and recognizing the target makes it hard for the end of the actuator to plan the trajectory. A 7 DOF visual servo mechanical arm simulation platform based on ROS and EtherCAT is designed and implemented. The CMVision algorithm is used on the ROS system to realize the fast identification and positioning while grabbing target. Based on the DH model of the manipulator arm to calculate,the inverse motion equation is solved to plan the trajectory of the manipulator arm. PID algorithm is applied to control. A real-time kernel is built on Linux. The end of the actuator can realize mobile control fast and accurately by EtherCAT technology. The test results show that the platform can achieve the goal of rapid identification and positioning and has good stability and operability.