Closed yiyongrobotics closed 4 years ago
@gouyiyong
1) We (our lab, past and now) never use 10Hz pose feedback to control the drone. If anyone (or paper) states 10 Hz, it's wrong. Also, I think we state in that paper we use 100 Hz pose feedback, clearly.
2) As I have explained, /vins_estimator/odometry is the pose in the sliding window of the visual-inertial estimator, but /vins_estimator/imu_propagate is with the IMU integration. 3) /vins_estimator/imu_propagate is accurate, very accurate, as long as your VINS runs properly and smoothly. 4) I think 10 Hz is far not good, based on my experience. And you can try it. Of course, you can use 10 Hz feedback if the performance satisfies your expectations.
@gouyiyong
- We (our lab, past and now) never use 10Hz pose feedback to control the drone. If anyone (or paper) states 10 Hz, it's wrong. Also, I think we state in that paper we use 100 Hz pose feedback, clearly.
- As I have explained, /vins_estimator/odometry is the pose in the sliding window of the visual-inertial estimator, but /vins_estimator/imu_propagate is with the IMU integration.
- /vins_estimator/imu_propagate is accurate, very accurate, as long as your VINS runs properly and smoothly.
- I think 10 Hz is far not good, based on my experience. And you can try it. Of course, you can use 10 Hz feedback if the performance satisfies your expectations.
@USTfgaoaa thanks for your detailed explanation, got it :)
are they /vins_estimator/odometry and /vins_estimator/imu_propagate much different? no any difference for control performance? and I wanna make sure if /vins_estimator/imu_propagate is accurate enough for the controller and the whole system. is it possible to run the controller with /vins_estimator/odometr 10 Hz? I read this paper https://onlinelibrary.wiley.com/doi/full/10.1002/rob.21732, you are using 10Hz odometry for the controller.