Skip to content

Latest commit

 

History

History
146 lines (117 loc) · 3.65 KB

real_robot.md

File metadata and controls

146 lines (117 loc) · 3.65 KB

导航+抓取demo准备工作

demo功能说明：车臂机器人在室内环境建图后，导航+抓取的工作流程：自主导航到A点，触发anygrasp抓取位姿估计后执行抓取操作，导航到B点后执行放置操作。

工作站：连接底盘、机械臂、夹爪、相机等硬件。 服务器：主要跑算法和应用,需要GPU。

注意：工作站和服务器在同一个局域网下，需要设置服务器ip地址与工作站在同一网段：192.168.6.xxx。

远程登录工作站

ssh [email protected]
密码：a

ros网络通信环境配置

工作站IP=192.168.6.102,假设服务器IP=192.168.6.103，以工作站为主机举例子：

主机设置

export ROS_HOSTNAME=192.168.6.102
export ROS_MASTER_URI=http://192.168.6.102:11311/

从机设置

export  HOSTNAME=192.168.6.103
export  ROS_MASTER_URI=http://192.168.6.102:11311/

建图

roslaunch dalu robot.launch
roslaunch dalu cartographer_mapping.launch use_sim_time:=false
roslaunch coffe_teleops keyboard_teleop.launch

#rosservice call /write_state "{filename: '<绝对路径存放地图信息>/<地图保存的名字>.pbstream'}"
rosservice call /write_state "{filename: '/home/maiya/maps/demo0425.pbstream'}"

导航

roslaunch dalu location.launch

可视化

rviz -d /home/maiya/demo_2d.rviz

注意：运行导航抓取之前利用键盘控制底盘旋转进行定位初始化。

roslaunch coffe_teleops keyboard_teleop.launch

导航+抓取

roslaunch dalu location.launch
roslaunch dalu GripperControl.launch
roslaunch visual_servo tNavArmGripperControl.launch

手眼标定

roslaunch visual_servo uHandEyeCalibrate.launch

标定结果保存在cali_result.json文件中。

其他

单独编译visual_servo功能包

catkin_make -DCATKIN_WHITELIST_PACKAGES="visual_servo"

启动anygrasp命令

cd ~/anaconda3/bin/ && source activate && conda activate anygrasp && cd /home/maiya/anygrasp_ws/src/anygrasp_sdk/grasp_live && demo_camera.sh

搭建房子demo启动节点服务

启动夹爪动作服务

rosrun robotiq_2f_gripper_control Robotiq2FGripperRtuNode.py /dev/ttyUSB1
# roslaunch robotiq_2f_gripper_action_server robotiq_2f_gripper_action_server.launch

动作客户端调用测试：
rosrun robotiq_2f_gripper_action_server robotiq_2f_gripper_action_server_client_test

启动UR3控制服务

roslaunch visual_servo tArmControlServer.launch

服务调用测试：
roslaunch visual_servo tArmControlClient.launch

真机实验

简单pick-place的demo

roslaunch visual_servo tNavArmGripperControl.launch
roslaunch visual_servo GripperControl.launch

启动AnyGrasp,即可对输出的pose执行动作

Just control UR3

roslaunch visual_servo uGraspControl.launch

Meanwhile, you need change codes at catkin_ws/src/visual_servo/utils/uGraspControl.cpp and /catkin_ws/src/visual_servo/include/visual_servo/UR3RtdeWrapper.h

复杂pick-place

master is robot:

robot:

roslaunch dalu localization.launch
roslaunch coffe_teleops keyboard_teleop.launch
roslaunch visual_servo GripperControl.launch
roslaunch visual_servo uGraspControl.launch

获得当前机器人所在位置：

rosrun tf tf_echo /map /base_link_calibration

server:

./demo_test.sh

robot:

________monitor rostopic via [rostopic echo /detect_grasps/pose_grasps]
rostopic  pub /grasp_trigger std_msgs/UInt8 "data: 1"