ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
2016-12-21 20:57:53 -0500 | commented answer | Yaw problem for IMU fusinon in robot_localization Thanks very much. @M@t So in your robot, if the GPS cannot work well, the trajectory will deviate a lot as well? Did you try @Tom Moore 's test bag in the r_l package? The IMU+ODOM fusion is much better than mine. |
2016-12-21 20:53:53 -0500 | received badge | ● Scholar (source) |
2016-12-21 20:53:52 -0500 | received badge | ● Scholar (source) |
2016-12-21 20:53:52 -0500 | received badge | ● Supporter (source) |
2016-12-21 16:20:29 -0500 | received badge | ● Student (source) |
2016-12-21 05:41:43 -0500 | received badge | ● Famous Question (source) |
2016-12-21 00:35:03 -0500 | received badge | ● Enthusiast |
2016-12-19 21:23:10 -0500 | commented question | Yaw problem for IMU fusinon in robot_localization @M@t Yes the IMU conforms to REP103 with ENU frame. I will prepare a small bag asap. |
2016-12-19 14:02:18 -0500 | received badge | ● Notable Question (source) |
2016-12-19 02:32:36 -0500 | commented question | robot_localization: map -> odom drifts in z-direction with zero-altitude set to true For your problem, you'd better try to make the two nodes as same setting in two-D mode parameter. I saw one is false and another is true. By the way, when you are fusing odom and imu, did your imu2 has an absolute yaw w.r.t. ENU frame? |
2016-12-19 01:52:03 -0500 | received badge | ● Editor (source) |
2016-12-16 05:19:25 -0500 | commented answer | How can I reduce drift in my robot's trajectory? I have similar sensors as your robot, I have a question. when you do the IMU and odom fusion, did your IMU has a yaw feedback ie. north or south? @M@t |
2016-12-07 19:37:15 -0500 | received badge | ● Popular Question (source) |
2016-11-29 22:59:01 -0500 | commented question | errors in using robot_localization package @krishna43 Hi I am doing the similar experiment. However, I found the /odom in TF is always produced by the bag file but not the ekf node. So that make the estimation and the odom the same. I saw your TF tree is correct. Any suggestions? |
2016-11-29 02:46:29 -0500 | commented question | Yaw problem for IMU fusinon in robot_localization @gvdhoorn Thanks for reminder. |
2016-11-29 01:46:52 -0500 | asked a question | Yaw problem for IMU fusinon in robot_localization EDIT1: The Yaw feedback from Mti-300 is a didital compass that in ENU frame. So it is not the same with /odom source. If I fusion data with IMU's Yaw, the result is very bad. If I only fusion with the yaw angular velocity, the result is acceptable. Is it correct? and Could anyone please tell me how to deal with the absolute yaw value? @M@t @Moderator EDIT2 I've got a GPS and a Mti-300 IMU producing IMU message at 400HZ. I am trying to use ekf_fusion to estimate the robot's position. Firstly, when I tried to use only one IMU & odom without the GPS, the result is very very bad. But I tried the testbag1, it works fine. Here is my simple launch file. The other launch file for the IMU node is omitted. I found the problem maybe caused by the yaw output from my IMU. Change the configuration matrix to The result is better,about 1 m error after 200m testing. However I know it is not corrct for using yaw from /odom. But it sems like I cannot use the yaw(from compass) from IMU. @Tom Moore Thnks very much, here are my sample messages. (more) |