Exercise 4

Overview

I learned using rqt_multiplot to visualize data and dealing with ros bag.

Demostration

exercise4.mp4

Instructions

• The node /ekf_localization uses measurements from wheel odometry, IMU sensor, and visual odometry to estimate the 3D robot pose by extended kalman filter

  # launch the file from exercise 3
  # in a new terminal
  rosnode info /ekf_localization
  • Publications:
    • /odometry/filtered [nav_msgs/Odometry]
  • Subscriptions:
    • imu/data [sensor_msgs/Imu]: 3D orientation
    • smb_velocity_controller/odom [nav_msgs/Odometry]: 2D pose from wheel

• Use rqt_multiplot to plot the path of simulated robot in xy-plane

• Install multiplot and run it

# install
sudo apt-get install ros-noetic-rqt-multiplot
# run
rqt_multiplot

• New configuration -> configure plot -> add curve -> save configuration as

• Download the rosbag smb_navigation.bag from the course website and inspect it

  rosbag info smb_navigation.bag

• Create a new launch file that starts the node ekf_localization_node

• Use rqt_multiplot to plot the path of the recorded robot

  # set the parameter `/use_sim_time` to true
  rosparam set use_sim_time true
  # play the bag file 
  rosbag play smb_navigation.bag --clock

  

• In the newly-created launch file (refer to the control.launch):

  • Add a robot_state_publisher node. See the reference here
  • Add a frame named smb_top_view. See here, section 6.3
  • /use_sim_time can be set true in launch file
  • Add rqt_multiplot
  • OPTIONAL 👏 Play the bag at 50% of the nominal speed