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

Odometry not correct

asked 2014-07-15 10:24:08 -0600

pnambiar gravatar image

I am trying to test the odometry of Segbot( With RMP 200 and Hokuyo) using the tests recommended in the ROS website. The translational test results looks reasonable, but the rotational results looks totally off. Any suggestions regarding what parameters needs to be changed to fix the problem?

edit retag flag offensive close merge delete

1 Answer

Sort by ยป oldest newest most voted
0

answered 2014-07-15 16:21:52 -0600

pwong gravatar image

If you're asking about the specific parameters for the segway, I'm not sure how the interface looks like for the segway rmp 200.

Suggestions.

Run your odometry node and subscribe to it with RVIZ. If you have a joystick or keyboard controls, see how transform moves as you control it with your joystick(Make sure the joystick drives your rmp correctly first). See if the joystick inputs match up to the transform of the odom tf(motion doesn't have to be exact, but just make sure if you flick the joystick left, the robot moves counter clock wise(etc).

I've used a segway rmp(although not exactly the 200 series), the input is in linear/angular velocity, correct? If so check to make sure a POSITIVE angular velocity input to the RMP makes it go counter clock wise, and a NEGATIVE angular velocity makes it go clock wise. I think I ran into this issue before where the RMP is opposite from the ROS standard. (Segway RMP goes cw with positive, ccw with negative, iirc).

Lastly, check your odometry math. The segway I used didn't output encoder values, so you have to find a way to derive it correctly given your output from the segway RMP.

Good luck with it!

edit flag offensive delete link more

Question Tools

1 follower

Stats

Asked: 2014-07-15 10:24:08 -0600

Seen: 450 times

Last updated: Jul 15 '14