Ask Your Question

localization problem when turning

asked 2018-07-23 04:34:28 -0600

kesuke gravatar image

updated 2018-08-11 00:02:46 -0600

jayess gravatar image

hi, i'm trying to use the navigation stack but i think that i have trouble with amcl , i have a two wheel mobile robot like a big version of turtlebot with 0.5 m radius equiped with Hokuyo laser , and since my wheel encoder gives a bad odometry to correct it i bought an imu sensor " Sparkfun Razor IMU 9DOF " and i used ekf localization to combine the odometry from encoder and the imu data , i had also remap the odom topic of the amcl to the new odometry topic provided by ekf
my problem is that sometimes the laserscan don't match with map specially when the robot stop , here is an example shown on the video you can see that when i'm rotating the robot the scan matches with the map but when i stop it there is a delay between scans and the map , i'm losing my mind i tried every thing i found on ros but it's still not working can you tell what's wrong with my configuration ?

here is the video : video

here is my amcl configuration :

  <arg name="use_map_topic"   default="false"/>
  <arg name="scan_topic"      default="scan"/>
  <arg name="initial_pose_x"  default="0.0"/>
  <arg name="initial_pose_y"  default="0.0"/>
  <arg name="initial_pose_a"  default="0.0"/>
  <arg name="odom_frame_id"   default="odom"/>
  <arg name="base_frame_id"   default="base_link"/>
  <arg name="global_frame_id" default="map"/>
  <node pkg="amcl" type="amcl" name="amcl">
    <param name="use_map_topic"             value="$(arg use_map_topic)"/>
    <remap from="/odom" to="/odometry/filtered"/>
    <!-- Publish scans from best pose at a max of 10 Hz -->
    <param name="odom_model_type"           value="diff"/>
    <param name="odom_alpha5"               value="0.1"/>
    <param name="gui_publish_rate"          value="10.0"/>
    <param name="laser_max_beams"             value="60"/>
    <param name="laser_max_range"           value="4.0"/>
    <param name="min_particles"             value="500"/>
    <param name="max_particles"             value="2000"/>
    <param name="kld_err"                   value="0.05"/>
    <param name="kld_z"                     value="0.99"/>
    <param name="odom_alpha1"               value="0.2"/>
    <param name="odom_alpha2"               value="0.2"/>
    <!-- translation std dev, m -->
    <param name="odom_alpha3"               value="0.2"/>
    <param name="odom_alpha4"               value="0.2"/>
    <param name="laser_z_hit"               value="0.5"/>
    <param name="laser_z_short"             value="0.05"/>
    <param name="laser_z_max"               value="0.05"/>
    <param name="laser_z_rand"              value="0.5"/>
    <param name="laser_sigma_hit"           value="0.2"/>
    <param name="laser_lambda_short"        value="0.1"/>
    <param name="laser_model_type"          value="likelihood_field"/>
    <!-- <param name="laser_model_type" value="beam"/> -->
    <param name="laser_likelihood_max_dist" value="2.0"/>
    <param name="update_min_d"              value="0.25"/>
    <param name="update_min_a"              value="0.2"/>
    <param name="odom_frame_id"             value="$(arg odom_frame_id)"/>
    <param name="base_frame_id"             value="$(arg base_frame_id)"/>
    <param name="global_frame_id"           value="$(arg global_frame_id)"/>
    <param name="resample_interval"         value="1"/>
    <!-- Increase tolerance because the computer can get quite busy -->
    <param name="transform_tolerance"       value="1.0"/>
    <param name="recovery_alpha_slow"       value="0.0"/>
    <param name="recovery_alpha_fast"       value="0.0"/>
    <param name="initial_pose_x"            value="$(arg initial_pose_x)"/>
    <param name="initial_pose_y"            value="$(arg initial_pose_y)"/>
    <param name="initial_pose_a"            value="$(arg initial_pose_a)"/>
    <remap from="scan"                      to="$(arg scan_topic)"/>

and here is the ekf yaml file i used :

frequency: 30
sensor_timeout: 0.1
two_d_mode: true
# Whether to broadcast the transformation ...
edit retag flag offensive close merge delete


What happens if you rotate the other direction? I suspect that the LIDAR isn't aligned correctly and you simply got lucky that it looks correct when rotating. I would expect the LIDAR to lag while rotating. Other direction probably much worse. Test and reply back.

billy gravatar image billy  ( 2018-07-23 18:10:45 -0600 )edit

thanks for answering , i tried what you said and it's the same result , i cheked that the LIDAR is aligned so i don't understand what's happennig it's like as if the amcl calculate the position late compared to the motion of the robot

kesuke gravatar image kesuke  ( 2018-07-24 03:08:38 -0600 )edit

Does this also happen without amcl? (Use odom as fixed frame in rviz)

Humpelstilzchen gravatar image Humpelstilzchen  ( 2018-07-24 03:11:20 -0600 )edit

when i put odom as fixed frame it's the same result , i know that the odometry i'm getting from encoder is poor that's why i added an imu but i dont know it's the odom that causes trouble

kesuke gravatar image kesuke  ( 2018-07-24 03:35:00 -0600 )edit

If I read your configuration file correctly you totally ignored the rotation from your wheels for odometry which is a good thing. So now we know the error comes from the IMU alone, correct? I suggest to calibrate the IMU and try again. Your problem might also come from magnetic interference.

Humpelstilzchen gravatar image Humpelstilzchen  ( 2018-07-24 03:40:58 -0600 )edit

Btw personally since I replaced my Razor IMU with a Bosch BNO055 based one (Tinkerforge IMU2) my navigation experience improved a lot.

Humpelstilzchen gravatar image Humpelstilzchen  ( 2018-07-24 03:47:49 -0600 )edit

@Humpelstilzchen i forget to calibrate it , i don't know if it will solve the problem but i will try it now

kesuke gravatar image kesuke  ( 2018-07-24 07:33:50 -0600 )edit

@kesuke have you solved this problem? I am facing something similar

aarontan gravatar image aarontan  ( 2018-07-27 10:26:28 -0600 )edit

2 Answers

Sort by ยป oldest newest most voted

answered 2018-08-13 11:42:59 -0600

Tom Moore gravatar image

I would only fuse velocities in your EKF, for starters. Stop fusing yaw from the IMU and X and Y position from your wheel odometry, and instead just fuse their velocities. BTW, you have roll and pitch enabled for your IMU, but you have two_d_mode on, so they'd be ignored anyway.

If I had to guess, your issue is one of covariances. You didn't include any sample wheel odometry or IMU messages (please do), but it looks to me that the EKF is trusting its own velocity estimate too much, so when the robot stops moving, it take the EKF a few cycles to converge to a 0 velocity. You can try increasing the process_noise_coviance for yaw velocity to see if that helps.

edit flag offensive delete link more

answered 2018-08-11 04:24:28 -0600

lalten gravatar image

To me it looks like amcl doesn't relocate because your rotation is < update_min_a ("Rotational movement required before performing a filter update"), so try reducing that parameter.

You should probably also increase the odom_alpha1/odom_alpha2 ("expected noise in odometry's rotation estimate") parameters if your IMU's orientation estimation is inaccurate.

Read through the parameter explanations in the wiki:

edit flag offensive delete link more


thank you i will try it

kesuke gravatar image kesuke  ( 2018-08-13 12:38:01 -0600 )edit

@kesuke did this solve the problem?

aarontan gravatar image aarontan  ( 2018-09-10 18:58:48 -0600 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Question Tools



Asked: 2018-07-23 04:34:28 -0600

Seen: 1,481 times

Last updated: Aug 13 '18