ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question
0

Exploding covariance in one state, rest is OK

asked 2022-08-04 03:06:33 -0500

Dats gravatar image

updated 2023-01-05 05:02:17 -0500

Tom Moore gravatar image

Hi, Im working on an AUV and trying to use robot_localization EKF for pose estimation. Im using a Water Linked A125 DVL, a pressure sensor and the internal IMU of the Pixhawk flight controller. My estimations are looking somewhat good in terms of movement in Rviz, but the covariance for X in the output from the EKF is exploding compared to all the other states. I cannot understand how X and Y can be so wildly different.

Static transformations between sensors are done in the launch file and are accurate withing +/- 1cm.

Does anyone have any idea what might be the cause of my troubles, or see anything wrong in my parameter file?

General info: ROS Melodic OS: Ubuntu 18.04

Params Yaml:

frequency: 15
sensor_timeout: 0.1
two_d_mode: false
transform_time_offset: 0.0
transform_timeout: 0.0
print_diagnostics: false
debug: false
publish_tf: true

publish_acceleration: false


# Frames

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


# Imu angular vel
imu0: /mavros/imu/data
imu0_config: [false, false, false,
              false, false, false,
              false, false, false,
              true, true, true,
              false,false,false]
imu0_nodelay: false
imu0_differential: false
imu0_relative: false
imu0_queue_size: 10
imu0_linear_acceleration_rejection_threshold: 4  #
imu0_remove_gravitational_acceleration: true


#DVL
odom0: /arrov/dvl/velocity
odom0_config: [false,  false, false, # x, y, z
               false, false, false,  # roll, pitch, yaw
               true, true, true,     # vx, vy, vz
               false, false, false,   # vroll, vpitch, vyaw
               false, false, false]  # ax, ay, az
odom0_queue_size: 10
odom0_nodelay: false
odom0_differential: false
odom0_relative: false
#odom0_pose_rejection_threshold: 5
#odom0_twist_rejection_threshold: 1

#Depth sensor - Z value
odom1: /arrov/odom
odom1_config: [false,  false, true, # x, y, z
               false, false, false,  # roll, pitch, yaw
               false, false, false,     # vx, vy, vz
               false, false, false,   # vroll, vpitch, vyaw
               false, false, false]  # ax, ay, az 
odom1_queue_size: 10
odom1_nodelay: false
odom1_differential: false
odom1_relative: false

initial_state: [ 0.0, 0.0, 0.0,
                 0.0, 0.0, 0.0,
                 0.0, 0.0, 0.0,
                 0.0, 0.0, 0.0,
                 0.0, 0.0, 0.0]

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,    0.05, 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,    0.01, 0,     0,     0,    0,    0,    0,    0,    0,    0,
                           0,    0,    0,    0,    0,    0,    0.1,   0,     0,    0,    0,    0,    0,    0,    0,
                           0,    0,    0,    0,    0,    0,    0,     0.1,   0,    0,    0,    0,    0,    0,    0,
                           0,    0,    0,    0,    0,    0,    0,     0,     0.1,  0,    0,    0,    0,    0,    0,
                           0,    0,    0,    0,    0,    0,    0,     0,     0,    0.1,  0,    0,    0,    0,    0,
                           0,    0,    0,    0,    0,    0,    0,     0,     0,    0 ...
(more)
edit retag flag offensive close merge delete

Comments

Please include a sample message from every sensor input.

Tom Moore gravatar image Tom Moore  ( 2022-09-26 04:17:56 -0500 )edit

Hi Tom, thanks for your reply. I've added examples of all the sensor messages I input to the EKF. I apologize for the bad formatting. As I am new to the forum I am not allowed to add images to my post. If there is anything else you need, please let me know. All help is truly appreciated.

Dats gravatar image Dats  ( 2022-09-29 09:47:24 -0500 )edit

1 Answer

Sort by ยป oldest newest most voted
1

answered 2023-01-05 05:08:24 -0500

Tom Moore gravatar image

Can you clarify "exploding"? You have no absolute pose data being fused for X (only velocities) so the error will grow without bound, but it should do so at a reasonable pace.

And I would expect X to grow at a faster rate than Y if you move in a straight line. Can you update your sensor messages so they have data from when the UUV is actually moving? I'm mostly interested in the DVL data. Also, what is your base_link -> dvl_link_child transform?

edit flag offensive delete link more

Question Tools

1 follower

Stats

Asked: 2022-08-04 03:06:33 -0500

Seen: 113 times

Last updated: Jan 05 '23