Lab3 » Historia » Wersja 6
Wersja 5 (Dawid Seredyński, 2018-06-13 17:22) → Wersja 6/8 (Dawid Seredyński, 2018-06-13 17:22)
h1. Lab 3: Forward kinematics
h2. Scope:
This laboratory consist of two parts:
* A short introduction.
* Writing your own ROS package.
h2. The task:
* Develop the system from the previous laboratory by either copying the package or modyfying it.
* Using C++ or Python create a ROS node NONKDL_DKIN that calculates forward kinematics of manipulator using analytical method (without KDL).
* The node gets current position of joints from joint_state_publisher node.
* The result (i.e. position of end effector of type geometry_msgs/PoseStamped) should be published on ROS topic and visualized in RVIZ.
* Do not interpolate movement.
* Please refer to RVIZ documentation http://wiki.ros.org/rviz/Tutorials
* Read parameters (e.g. lengths of links) from ROS parameter server.
!https://www.robotyka.ia.pw.edu.pl/redmine/attachments/download/203/lab3-scheme-a.png(graph a: ROS nodes for ANRO laboratory 3)!
* Create a second node KDL_DKIN that uses KDL, with the same functionality as NONKDL_DKIN.
* Connect it to existing nodes in place of NONKDL_DKIN.
!https://www.robotyka.ia.pw.edu.pl/redmine/attachments/download/204/lab3-scheme-b.png(graph b: ROS nodes for ANRO laboratory 3)!
* Test the results in RVIZ. Compare nodes NONKDL_DKIN and KDL_DKIN with data from robot_state_publisher.
* Check if the calculated pose is the same as the pose calculated by robot_state_publisher. Compare the poses in RVIZ.
* All three nodes that calculate forward kinematics should publish to different topics.
* Reference frames of the end effector should coincide. Set different visualization parameters (e.g. width, length of arrows) to see them all.
* The documentation of tests should be supplemented with plots of end effector position calculated using different methods and plots of configuration space position.
!https://www.robotyka.ia.pw.edu.pl/redmine/attachments/download/205/lab3-scheme-c.png(graph c: ROS nodes for ANRO laboratory 3)!
* Joint limist should be taken into account.
* If joint position is outside limits, the pose of end effector should not be published and error should be published to ROS log.
* Add wiki page to your repository (the documentation). The page should contain description of source files, instruction for launching and description of tests.
* Use separate roslaunch files for:
* roscore and RVIZ (it should be running until URDF file is modified).
* The rest of the system.
* Test the system and show it to the tutor.
h2. Scope:
This laboratory consist of two parts:
* A short introduction.
* Writing your own ROS package.
h2. The task:
* Develop the system from the previous laboratory by either copying the package or modyfying it.
* Using C++ or Python create a ROS node NONKDL_DKIN that calculates forward kinematics of manipulator using analytical method (without KDL).
* The node gets current position of joints from joint_state_publisher node.
* The result (i.e. position of end effector of type geometry_msgs/PoseStamped) should be published on ROS topic and visualized in RVIZ.
* Do not interpolate movement.
* Please refer to RVIZ documentation http://wiki.ros.org/rviz/Tutorials
* Read parameters (e.g. lengths of links) from ROS parameter server.
!https://www.robotyka.ia.pw.edu.pl/redmine/attachments/download/203/lab3-scheme-a.png(graph a: ROS nodes for ANRO laboratory 3)!
* Create a second node KDL_DKIN that uses KDL, with the same functionality as NONKDL_DKIN.
* Connect it to existing nodes in place of NONKDL_DKIN.
!https://www.robotyka.ia.pw.edu.pl/redmine/attachments/download/204/lab3-scheme-b.png(graph b: ROS nodes for ANRO laboratory 3)!
* Test the results in RVIZ. Compare nodes NONKDL_DKIN and KDL_DKIN with data from robot_state_publisher.
* Check if the calculated pose is the same as the pose calculated by robot_state_publisher. Compare the poses in RVIZ.
* All three nodes that calculate forward kinematics should publish to different topics.
* Reference frames of the end effector should coincide. Set different visualization parameters (e.g. width, length of arrows) to see them all.
* The documentation of tests should be supplemented with plots of end effector position calculated using different methods and plots of configuration space position.
!https://www.robotyka.ia.pw.edu.pl/redmine/attachments/download/205/lab3-scheme-c.png(graph c: ROS nodes for ANRO laboratory 3)!
* Joint limist should be taken into account.
* If joint position is outside limits, the pose of end effector should not be published and error should be published to ROS log.
* Add wiki page to your repository (the documentation). The page should contain description of source files, instruction for launching and description of tests.
* Use separate roslaunch files for:
* roscore and RVIZ (it should be running until URDF file is modified).
* The rest of the system.
* Test the system and show it to the tutor.