Robot Localization with Odometry and UM7
Hello all,
I am trying to get localization to work for a custom car like robot.
I am able to publish odom from wheel encoders and imu/data from my UM7 imu sensor.
I would like to fuse together these two data using the robot localization package and then show it with baselink in rviz. Without proper localization of baselink, I will not be able to run the navigation stack properly.
In my current setup, rviz is able to show forward and backward movement with base_link; however, when I manually rotate the um7 imu sensor, it does not reflect within rviz. I can see that the odometry/filtered topic is receiving input from the um7 but how do i visualize this in rviz?
My end goal is to place this um7 on my custom car like robot. I want rviz to display the base_link mimicking its real life counterpart in terms of position/orientation and velocities (moving forward/backward, turing left/right...etc). It is to my understanding that this is what the robot localization package is for, please correct my understanding if i am making a mistake.
Here are two videos describing my problem:
1st video shows the screen recording from my computer. You can see that moving forward and backward is okay, but I do not know how to turn the base_link.
2nd video shows what I am trying to do
Below is the configuration file for the robot_localization package:
frequency: 50
sensor_timeout: 0.1
two_d_mode: true
transform_time_offset: 0.0
transform_timeout: 0.0
print_diagnostics: true
debug: false
debug_out_file: /path/to/debug/file.txt
publish_tf: true
publish_acceleration: false
odom_frame: odom # Defaults to "odom" if unspecified
base_link_frame: base_link # Defaults to "base_link" if unspecified
world_frame: odom # Defaults to the value of odom_frame if unspecified
odom0: /odom
odom0_config: [false, false, false,
false, false, true,
true, true, false,
false, false, true,
false, false, false]
odom0_queue_size: 2
odom0_nodelay: false
odom0_differential: false
odom0_relative: false
odom0_pose_rejection_threshold: 5
odom0_twist_rejection_threshold: 1
imu0: /imu/data
imu0_config: [false, false, false,
false, false, true,
false, false, false,
false, false, true,
false, false, false]
imu0_nodelay: false
imu0_differential: false
imu0_relative: true
imu0_queue_size: 5
imu0_pose_rejection_threshold: 0.8 # Note the difference in parameter names
imu0_twist_rejection_threshold: 0.8 #
imu0_linear_acceleration_rejection_threshold: 0.8 #
imu0_remove_gravitational_acceleration: true
use_control: true
stamped_control: false
control_timeout: 0.2
control_config: [true, false, false, false, false, true]
acceleration_limits: [1.3, 0.0, 0.0, 0.0, 0.0, 3.4]
deceleration_limits: [1.3, 0.0, 0.0, 0.0, 0.0, 4.5]
acceleration_gains: [0.8, 0.0, 0.0, 0.0, 0.0, 0.9]
deceleration_gains: [1.0, 0.0, 0.0, 0.0, 0.0, 1.0]
process_noise_covariance: [0.05, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0.05, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0.06, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0.03, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0.03, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0.06, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0.025, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0.025, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0.04, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0.01, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.01, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.02, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.01, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.01, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.015]
initial_estimate_covariance: [1e-9, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 1e-9, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 1e-9, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 1e-9, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 1e-9, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 1e-9, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 1e-9, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 1e-9, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 1e-9, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 1e-9, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1e-9, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1e-9, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1e-9, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1e-9, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1e-9]
Asked by aarontan on 2018-09-10 18:34:13 UTC
Answers
A couple things.
- Please get rid of all the advanced parameters, such as the rejection thresholds. Also turn of
use_control. Those really are meant for advanced usage. What I typically recommend is that you start with a single sensor and a simple config, then work your way up. - Looking at your configuration, you are fusing yaw as an absolute pose variable from your odometry and from the IMU. So let's say your wheel odometry says you have a yaw of 0. Then you pick up your IMU and rotate it. Now you have one data source that reads 0, and another that reads non-zero. I would expect the filter to settle on a value in between them, but it really depends on your covariances. You may also have an issue with your IMU data.
You didn't include sample sensor messages, so I can't comment on whether your IMU and wheel encoder data need any changes, but certainly issue (2) above would be a big problem.
Asked by Tom Moore on 2018-10-24 10:11:52 UTC
Comments
Have you calibrated the IMU? Calibrating individual sensors (accelerometer, gyroscope, magnetometer) is important for proper functioning of the IMU.
Asked by Subodh Malgonde on 2018-09-11 00:53:49 UTC
Also before you fuse the IMU data with robot_localization can you visualize IMU attitude and heading, something like this https://user-images.githubusercontent.com/5463437/43065886-c729e4c2-8e80-11e8-8d47-905032444590.gif. I am using the Razor IMU and its github repo comes with a visualization tool.
Asked by Subodh Malgonde on 2018-09-11 00:55:08 UTC
@Subodh Malgonde yes I have calibrated it and I can see the imu/data published. The UM7 does not have a visualization tool like the razor. In your application, when you fuse together odometry with imu data, are you able to complete what I stated as my end goal? (base_link motion mimicking real lif
Asked by aarontan on 2018-09-11 08:25:43 UTC
Yes I am able to do that. One suggestion, before you try it on the actual hardware you should try simulating a wheel encoder + IMU in Gazebo, by setting sensor noise to zero. That way you will know if something is wrong with your robot_localization config.
Asked by Subodh Malgonde on 2018-09-12 00:04:04 UTC
@aarontan could you help me by guiding me how were you able to generate odometry data from wheel encoders?
Asked by Raza Jafri on 2020-08-12 03:01:28 UTC