robot_localization: odometry/filtered going to zero when using IMU
Hello,
I am trying to use the ekflocalizationnode to fuse odom and IMU data with a simulated robot in gazebo. In my simulation the robot starts in a position different from (0,0,0), something like (18, 2, 0). Therefore, the /odom topic is also (18, 2, 0) when the simulation starts. When using the ekf_localization node if I use only odom the filtered odom seems to be working, but when I add the IMU the odom/filtered starts at (0,0,0) which is wrong in this case, how can I set up the IMU to make this work?
The ekf params that I am using:
frequency: 30
silent_tf_failure: false
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
map_frame: map # Defaults to "map" if unspecified
odom_frame: odom # Defaults to "odom" if unspecified
base_link_frame: base_footprint # Defaults to "base_link" if unspecified
world_frame: odom # Defaults to the value of odom_frame if unspecified
odom0: /odom
odom0_config: [true, true, false,
false, false, false,
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_pub
imu0_config: [false, false, false,
true, true, true,
false, false, false,
true, true, true,
true, true, false]
imu0_nodelay: false
imu0_differential: false
imu0_relative: false
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: false
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 rezenders on 2020-12-07 08:36:59 UTC
Answers
Do you need your EKF coordinates to match the coordinate frame of Gazebo?
The issue is just one of timing. If you don't supply an initial state, the first measurement you receive is used to initialise the filter. When you start with just odom data, the first measurement the filter gets puts the state estimate where Gazebo says it is.
When you add the IMU, my guess is the filter receives an IMU measurement before an odometry measurement. Since the IMU doesn't have pose data, it initialises the filter to a (0, 0) position, but with the given orientation.
Anyway, once you receive some odometry measurements, the filter should still move to the odometry pose, though it may take a few measurements.
If it were me, I'd not worry about whether the coordinate frame from Gazebo matches the EKF one. I'd just fuse wheel encoder and IMU velocity data. The filter will always start at the origin, but it'll be consistent.
Asked by Tom Moore on 2020-12-09 06:04:30 UTC
Comments
Actually it is standard practice for the odom frame to be initialized where the robot starts. So I think it is actually the first case, where the robot odom starts at (18,2) that is mistaken
Asked by JackB on 2020-12-08 12:11:58 UTC
I spawned my model using the gazebo_ros spawn_model node passing the arguments with the coordinates I wanted, do you know if I should do it in a different way? But I wonder if the odom frame was at 0,0,0 for my robot, what would happen if I spawned another one in a different position? Wouldn't it be dislocated for the 2nd robot? I guess the odom is in a fixed position in gazebo to prevent this?
Asked by rezenders on 2020-12-08 16:59:52 UTC