Search Results

In your navsat_transform: You say you are using the differentiated GPS values to generate a yaw that navsat_transform_node uses as its IMU heading, correct? … Next, get the map frame EKF instance working without GPS data. I tend to fuse the same sources for that instance, with the same parameters. Finally, add the GPS data. …

, false, true, false, false, false, false, false, true, false, false, false] You can see that you are fusing absolute X and Y position from your GPS …

Our IMU yaw rate is obtained from our vehicle's CAN data of yaw rate, and we integrate that value. We don't have a direct IMU source. Sorry, are you saying that you don't have a magnetometer that gi …

You should include your full EKF and navsat_transform_node config, a sample input message from every sensor input, and the value of any static transforms. navsat_transform_node will account for the GPS …

This is unrelated to your issue, but your magnetic_declination_radians has a comma instead of a decimal point: <param name="magnetic_declination_radians" value="0,1689479"/> What happens if you don' …

If it's (2), I would expect that, because GPS will drift, and your GPS coordinates when you start between any two runs will change. … In what frame should I give move_base the points if I want to make a GPS waypoint follower? In your case, you'd want to issue map-frame coordinates. …

That's a bug. This line has the cartesian_x and cartesian_y parameters backwards (cartesian_y should come first). Please file a ticket and I'll fix it. Thank you! Originally posted by Tom Moore with …

This has caused /odometry/filtered and odometry/gps to give much better results. …

Here's how it should look: EKF 1 (odom frame) Inputs: /imu/data, /odom Outputs: /odometry/filtered, odom->base_link transform EKF 2 (map frame) Inputs: /imu/data, /odom, /odometry/gps Outputs: odometry … /filtered_gps, map->odom transform navsat_transform_node Inputs: /imu/data/gps, /ublox_gps_rover/fix, odometry/filtered_gps <-- You need this Outputs: /odometry/gps Your issue (or at least one issue that …

only needs one IMU message to generate the required transform, so one option is to make a node that waits until your robot has driven X meters, and then uses the relative poses of the start and current GPS …

Offhand, I note three things: You aren't fusing yaw or yaw velocity from your wheel odometry data. That's going to cause a covariance explosion in the output, which will have all kinds of ugly side …

Let's say you have three data sources: wheel odometry, IMU data, and GPS. You have two EKFs. The first one fuses just wheel odometry and IMU data. The second fuses wheel odometry, IMU, and GPS data. … Final note: if all you have is an IMU and GPS, then (a) this isn't very well supported by r_l, and (b) there's no reason for you to use two EKFs. …

r_l is going to use that time stamp, and if the clocks are wildly out of sync, it could definitely have strange effects on your state estimate. At most, if you're going to submit a PR to nmea_navsat_ …

It's all really just a reflection of what's specified in REP-105. From the "Frame Authorities" section: The transform from odom to base_link is computed and broadcast by one of the odometry sources. …

This is what happens between your GPS measurements. Then you receive a GPS measurement, and the filter does a weighted average between that GPS data and your current state estimate. … Even then, the GPS is very likely to violate the covariance it reports, so your motion won't always be smooth. …