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

4-wheel robot localization in global

asked 2015-04-07 14:02:07 -0500

crazymumu gravatar image

updated 2015-04-07 14:05:20 -0500

image description I have odometry_publish node to subscribe velocity from vfh_node with cmd_vel topic; publish the tf between robot between global original point;

And target_tf_node subscribe tf from odometry_publish node to compute target position into robot's frame and publish target position(robot frame) with topic target_theta.
---------this step is localization, because get tf to know where the robot in global frame so that transform target into robot frame.

vfh_node subscribe target position from target_tf_node and obstacle information from laser sensor node, manipulate those data to determine a velocity publish through cmd_vel topic which drive the robot in gazebo.

I just want to know is that idea ok? or just tf information is not enough for localization and lead to big error in robot's location.

edit retag flag offensive close merge delete

1 Answer

Sort by ยป oldest newest most voted

answered 2022-06-28 09:03:21 -0500

vonunwerth gravatar image

Generally 4-wheel-robot localization only based on the odometry will lead to big errors. Actually, you have two too many wheels, which you take along "only for stability", to put it somewhat exaggeratedly.

This redundancy leads to this "uncontrolable" form of skid steering, which generates some slippage with your wheels. Since you can't track exactly how much the wheels have slid over somewhere, the more you drive, the more your error will increase.

To get better results, you could combine sensor data, like odometry, imu, ... with the robot_localization package. You could also think about generating a map based on your laser scans and integrate that.

edit flag offensive delete link more

Question Tools

1 follower


Asked: 2015-04-07 14:02:07 -0500

Seen: 521 times

Last updated: Jun 28 '22