fly316/PX4-Autopilot
7
0
Fork 0
mirror of https://gitee.com/mirrors_PX4/PX4-Autopilot.git synced 2026-07-06 06:40:35 +08:00
Code Issues Packages Projects Releases Wiki Activity
33,107 Commits 852 Branches 161 Tags
20c2fe6d285867b51ec68a5199391385cf524e1e
Commit Graph

435 Commits

Author SHA1 Message Date
Daniel Agar 20c2fe6d28 estimator messages add explicit timestamp_sample 
- timestamp is uORB publication metadata
 - this allows us to see what the system saw at publication time plus the latency in estimation
 2020-09-04 10:48:26 -04:00
Daniel Agar 9215fb57a2 ekf2: advertise immediately to ensure consistent uORB instance numbering 2020-09-04 10:48:26 -04:00
Daniel Agar 9ccc1db649 estimator_status split out estimator_states 2020-09-04 10:48:26 -04:00
Daniel Agar 33a7fed240 ekf2: split out header and rename 2020-09-04 10:48:26 -04:00
Daniel Agar 4879a4e2ef ekf2: use single uORB::Subscription, but change instance 2020-09-03 15:41:33 -04:00
Daniel Agar 6ff361479c uORB: introduce SubscriptionMultiArray for working with multi-instances 2020-09-02 12:46:47 -04:00
Hamish Willee 979243f38f params: make parameter units more consistent (#15502) 2020-08-24 11:33:08 +02:00
Daniel Agar 971b1e4b4d sensors: move mag aggregation to new VehicleMagnetometer WorkItem 
- purge all reminaing magnetometer IOCTL usage
 - mag calibration add off diagonal (soft iron) scale factors
 2020-08-21 10:12:13 -04:00
Daniel Agar 97fc1db768 vehicle_local_position: rename yaw -> heading and add reset logic 
- vehicle_global_position yaw removed (redundant)
 2020-08-10 11:42:03 +02:00
Daniel Agar fa44cd5cd5 px4_work_queue: rename navigation_and_controllers -> nav_and_controllers 
- keep the thread within the current NuttX max length
 2020-07-14 09:48:47 -04:00
dominiknatter d8831c5133 ekf2: Ignore optical flow samples with too large integration time spans. Fixes #14165 
Within ekf2, optical flow messages (amongst others) are fused to the state estimates. It might occur that optical flow sensors report unreliable and unrealistic spikes. In that case, the state estimator went crazy so far and just ignored optical flow values from that moment on.

The common thread for all these spikes seems to be a too high integration time span. Therefore, this fix adds a simple logic that ignores unrealistically high integration time spans. As a threshold, 1 second was chosen.

Reported-by: Dominik Natter <dominik.natter@gmail.com>
 2020-06-29 12:00:29 -04:00
Beat Küng 71dcf8d619 sitl: use lockstep components API 
- avoids the need for ekf2_timestamp publications by q and lpe
- adds logger to the lockstep cycle and makes it poll on ekf2_timestamps
  or vehicle_attitude. This avoids dropped samples (required for replay).
 2020-06-23 11:53:55 -04:00
Daniel Agar 11585dfb67 ekf2: decrease default GPS horizontal velocity noise 2020-06-02 08:18:12 -04:00
Daniel Agar e34bdb4be9 move IMU integration out of drivers to sensors hub to handle accel/gyro sync 
- IMU integration move from drivers (PX4Accelerometer/PX4Gyroscope) to sensors/vehicle_imu 
 - sensors: voted_sensors_update now consumes vehicle_imu
 - delete sensor_accel_integrated, sensor_gyro_integrated
 - merge sensor_accel_status/sensor_gyro_status into vehicle_imu_status
 - sensors status output minor improvements (ordering, whitespace, show selected sensor device id and instance)
 2020-05-30 11:07:54 -04:00
Daniel Agar 0ec48cfef3 ekf2: increase default baro noise 2 -> 3.5 m 2020-05-28 08:35:07 -04:00
kritz 3897030c6f Support odometry velocity in body and local frame (#14703) 
* Update submodule ECL

* increase lower bound on EVV param
 2020-05-13 12:43:02 +02:00
Mohammed Kabir 5ffe88672e vehicle_odometry: add timestamp_sample field for latency monitoring 2020-04-28 13:58:43 -04:00
kritz ccaa103164 ekf2: Stop getting velocity variance from pose covariance matrix (#14779) 2020-04-28 10:05:30 -04:00
Daniel Agar 326d8efc16 move attitude controllers to new wq:attitude_ctrl 2020-04-27 21:34:35 -04:00
Hamish Willee 8236b8da81 Fix up doc links to point to master branch 2020-04-21 08:58:40 +02:00
Daniel Agar cf37be8c44 ekf2 handle accelerometer clipping 
- track clipping per IMU axis and pass through to ecl/EKF
 - update ecl/EKF to include delta velocity clipping changes (PX4/ecl#663)
 2020-04-07 20:11:08 -04:00
kamilritz a70aa6349a ECL interface: output vector quantities by return-by-value 2020-04-01 11:47:39 +02:00
Hamish Willee 1288db9314 ekf2_params.c: EKF2_MAG_TYPE: Info about yaw without mag 2020-03-25 09:41:54 +01:00
bresch 0f2f72a757 Ekf2: Add bitmask to select terrain estimator aiding sources 2020-03-19 20:27:39 +03:00
CarlOlsson 58ded8d4c0 ekf2: remove unused function declatation 2020-03-18 14:33:11 +01:00
Paul Riseborough ab92b46e69 Update ecl to add ability to recover from bad magnetic yaw 
* msg: Add EKF-GSF yaw estimator logging data
* ecl: update to version with EKF-GSF yaw estimator
* ekf2: Add param control and logging for EKF-GSF yaw estimator
* logger: Add logging for EKF-GSF yaw esimtator
 2020-03-15 12:57:25 -04:00
Daniel Agar 0eca583ecf sensors: move baro aggregation to new sensors/vehicle_air_data 2020-03-12 19:06:34 -04:00
Daniel Agar a89b69b0ea vehicle_global_position: remove velocity fields (duplicates of local vx, vy, vz) 
* attitude_estimator_q: get velocity from local position (if available)
 2020-03-11 23:57:43 -04:00
RomanBapst 4ddf0f9dc1 ekf2_main.cpp: fixed bug in the calculation of the mag bias calibration time 
Signed-off-by: RomanBapst <bapstroman@gmail.com>
 2020-03-10 08:54:04 +01:00
CarlOlsson 8de1f5229b EKF2: use fixed time constant for GPS blending 
Signed-off-by: CarlOlsson <carlolsson.co@gmail.com>
 2020-03-04 16:46:20 +01:00
CarlOlsson 357700aa8d EKF2 GPS blending: fall out of blending if one module lose 3D fix 
Signed-off-by: CarlOlsson <carlolsson.co@gmail.com>
 2020-03-04 16:46:20 +01:00
CarlOlsson a10a228092 EKF2 GPS blending: reset relative position offsets if blending not feasible 
Signed-off-by: CarlOlsson <carlolsson.co@gmail.com>
 2020-03-04 16:46:20 +01:00
CarlOlsson 7e00dcd382 EKF2: Fix GPS blending time jitter 2020-02-26 07:39:47 +01:00
CarlOlsson a61f1647ad replay: remove GPS relative timestamps 
Signed-off-by: CarlOlsson <carlolsson.co@gmail.com>
 2020-02-07 08:19:22 +01:00
CarlOlsson 37db7d3bba ekf2: define max number of range finders seperate from max number of mutli uORB topics 
Signed-off-by: CarlOlsson <carlolsson.co@gmail.com>
 2020-02-05 08:03:23 +01:00
Hamish Willee 9886b11f99 EKF params - clarify body frame origin/orientation 2020-01-30 07:29:11 -05:00
Daniel Agar 10410fc868 msg: rename sensor_bias -> estimator_sensor_bias 2020-01-27 09:03:27 +01:00
kamilritz 5a5b3a44ee Update flow interface 2020-01-24 16:49:17 +03:00
kritz 47cab3dba5 Update external vision interface (#13989) 
* Update external vision interface

* Update ECL Version
 2020-01-23 15:36:08 +01:00
kamilritz 8d25dcd543 Update EKF auxiliary velocity interface 2020-01-23 13:27:30 +01:00
kamilritz 9d184ea3d6 Update EKF range finder interface 2020-01-23 13:27:30 +01:00
kamilritz 230d7f2558 Update EKF airspeed interface 2020-01-23 13:27:30 +01:00
kamilritz 8772e09963 Update EKF gps interface 2020-01-23 13:27:30 +01:00
kamilritz 19b7421f86 Update EKF baro interface 2020-01-23 13:27:30 +01:00
kamilritz 439a20bfcb Update EKF mag interface 2020-01-23 13:27:30 +01:00
Daniel Agar 697dbfb9f8 sensors/vehicle_imu: incremental step towards multi-EKF 2020-01-22 18:04:29 -05:00
Daniel Agar 21a8d7db7f WorkItem modules: Run() shouldn't be public 2020-01-22 12:03:03 -05:00
kamilritz 04e09d772e Add missing mag calibration biases 2020-01-21 13:36:08 +03:00
kamilritz a3d2254e44 Move mag downsampling to ECL lib 2020-01-21 13:36:08 +03:00
kamilritz 4e8665082e Move baro downsampling and dynamic pressure comp to ECL lib 2020-01-21 13:36:08 +03:00