Commit Graph

39790 Commits

Author SHA1 Message Date
Daniel Agar 7cb8ed3a1f sensors/vehicle_gps_position: don't work with 64 time with 32 bit floats 2022-01-21 09:18:49 -05:00
JaeyoungLim 0607982b23 Publish orbit status also when the waypoint is a loiter waypoint (#19048)
* Publish orbit status also when the waypoint is a loiter waypoint

* Remove redundant orbit status publishing
2022-01-21 09:13:03 +01:00
Beat Küng bcd057ac3e uORB: fix copy-paste mistake in orb_print_message_internal
Could have led to invalid memory access.
2022-01-20 12:58:20 -05:00
bresch a2260e53da ekf2: replace yaw failure detection criteria
The existing logic using the angle between velocity vectors failed to
determine a yaw failure in practice because the state velocity is often
too small compared to its uncertainty to be used. In all the failures
reported, the yaw emergency estimator converged properly and yaw reset
would have fixed the issue.
A much simpler check using the yaw difference between the main EKF
and the emergency estimator is now used to tell if the vel/pos update
failure is most likely caused by a wrong heading.
2022-01-20 11:54:24 +01:00
Daniel Agar 446729566d platforms/nuttx: px4io_serial always cleanup DMA before next bus exchange
- this really shouldn't be necessary, but worst case it's harmless and
much better than potentially falling out of the sky
2022-01-19 18:00:18 -05:00
Jaeyoung Lim 0dea56f88b Use current_sp for publishing orbit status 2022-01-19 12:00:03 -05:00
Jaeyoung-Lim dec5e7e9c8 Publish orbit status whenever the vehicle is in loiter
This commit moves the publish orbit status to be published whenever the vehicle is considering a loiter setpoint
2022-01-19 12:00:03 -05:00
Thomas Stastny 19c98b8841 npfg: fix typo in Vector2f indexing 2022-01-19 11:17:44 -05:00
bresch e89e3c1b0c ekf2: extract logic to test yaw emergency estimate quality 2022-01-19 09:05:41 -05:00
bresch ea80c5027e ekf2: split yaw estimator state getter into several functions 2022-01-19 09:05:41 -05:00
dagar 476c1e5c09 [AUTO COMMIT] update change indication 2022-01-19 09:04:53 -05:00
bresch ffebd8e771 ekf2: initialize cos tilt variable assuming no tilt 2022-01-19 09:04:53 -05:00
Daniel Agar 92a5bbe97f ekf2: improve ring buffer sizing and dynamically allocate 2022-01-19 09:04:53 -05:00
Daniel Agar 36605bfff6 sensors: lower SENS_MAG_RATE default
- reduce estimator 3D mag fusion cpu load (-3% on Pixhawk4)
2022-01-19 08:51:57 -05:00
Matthias Grob c9f7c20d46 MulticoperPositionController: time failsafe messages together
The following output instead of printing the
action "stop and wait" just once:

WARN  [mc_pos_control] invalid setpoints
WARN  [mc_pos_control] invalid setpoints
WARN  [mc_pos_control] Failsafe: stop and wait
WARN  [mc_pos_control] invalid setpoints
WARN  [mc_pos_control] Failsafe: stop and wait
WARN  [mc_pos_control] invalid setpoints
WARN  [mc_pos_control] Failsafe: stop and wait
2022-01-19 14:31:19 +01:00
Matthias Grob d4e356a1ac PosititionControl: fix integrator windup with invalid setpoint 2022-01-19 14:31:19 +01:00
Matthias Grob 8811482f1d PositionControlTest: add integrator windup with invalid setpoint combination unit test
This bug was by chance found during simulation testing and debugging.
The unit test is to easily reproduce and cover this case.
2022-01-19 14:31:19 +01:00
Silvan Fuhrer 632dfa55e6 FW Pos C: add option to disable airspeed setpoint via stick input
-rename FW_POSCTL_INV_ST to FW_POS_STK_CONF and make bitmask out of it:
  - bit 0: alternative mapping (height rate on throttle stick, airspeed on pitch)
  - bit 1: enable/disable airspeed setpoints via stick

Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2022-01-19 10:13:07 +01:00
Silvan Fuhrer cbb8c90245 ControlAllocation: fix calculation of roll/pitch normalization scale
Take into account the actual number of roll and pitch acutators,
instead of assuming that all actuators have a roll/pitch component.

Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2022-01-19 08:22:12 +01:00
Silvan Fuhrer 0950bb81ab ControlAllocation: update normalization scale only if matrix updated is forced
The forced flag is used to distinguish between updates due to a configuration
(parameter) change (only enabled when disarmed), and matrix updates due to motor
tilt change. Only update the normalization scale if the forced flag is true, and
use a tilt angle of vertical position for it to have the scales tilt-invariant.

Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2022-01-19 08:22:12 +01:00
CUAVmengxiao 258a563dd5 barometer: Add ICP10100 and ICP1011 2022-01-18 19:14:40 -05:00
FARHANG fe44e281e5 ROMFS: Holybro X500 v2 airframe addition 2022-01-18 15:11:41 -05:00
Daniel Agar dd6b7fa98f sensors: switch status PX4_INFO -> PX4_INFO_RAW 2022-01-18 15:00:18 -05:00
Daniel Agar b1d2a0cc4e sensors: simplify timestamp data validator handling 2022-01-18 14:57:18 -05:00
Daniel Agar 8067207ea6 px4_work_queue: rename serial port WQs 2022-01-18 14:56:15 -05:00
bresch 953c90d3a6 failure detector: change imbalanced propeller metric
Use standard deviations instead of variances to get a linear metric
2022-01-18 12:24:35 -05:00
bresch 408cf011b2 failure detector: fix dt computation 2022-01-18 12:24:35 -05:00
Daniel Agar 7eee949a81 Jenkins: hardware tolerate irrelevant uorb status failure 2022-01-18 12:23:05 -05:00
Daniel Agar 548b7d5ece sensors/vehicle_imu: on calibration change don't resume in flight calibration learning immediately 2022-01-18 11:22:41 -05:00
Daniel Agar ece09064c4 parameters: perform verification pass twice
- on NuttX with MTD storage this is to ensure we are verifying what's
actually stored on ramtron or eeprom, not just what's cached by bchlib
2022-01-18 09:47:28 -05:00
alexklimaj cdb6a437a0 Fix UAVCAN beep not started 2022-01-17 21:49:53 -05:00
PX4 BuildBot 0bb9cce1ce Update submodule mavlink to latest Tue Jan 18 00:39:07 UTC 2022
- mavlink in PX4/Firmware (37e04f56da3916dcb518eeba66b0888db1cd8c32): https://github.com/mavlink/mavlink/commit/75204adad281e017e58e0a68fdb0a478c2b74d6f
    - mavlink current upstream: https://github.com/mavlink/mavlink/commit/3d8092006ec1b2321f74e84039dadb1ebd3d1f37
    - Changes: https://github.com/mavlink/mavlink/compare/75204adad281e017e58e0a68fdb0a478c2b74d6f...3d8092006ec1b2321f74e84039dadb1ebd3d1f37

    3d809200 2022-01-14 Hamish Willee - MAV_PROTOCOL_CAPABILITY_FLIGHT_INFORMATION - mark as reserved capability (#1775)
68ff649f 2022-01-13 Hamish Willee - MAV_PROTOCOL_CAPABILITY_MISSION_FLOAT - deprecate (#1768)
ae531d97 2022-01-13 Peter Barker - Move "external" dialects to be normal dialects (#1761)
00a3f8ed 2022-01-12 Julian Oes - Revert "Delete unused cmake tree (#1760)" (#1782)
aa0764bd 2022-01-12 Hamish Willee - MAV_PROTOCOL_CAPABILITY_PARAM_UNION - rename/reserved (#1771)
c3b46620 2022-01-12 Hamish Willee - Update mavlink_udp.c example readme to use mavlink 2 (#1780)
0449a5ec 2022-01-12 Tatsuya Yamaguchi - add extended sensor status bits to MAVLink UDP example (#1778)
fc3539e1 2022-01-10 Josh Henderson - ardupilotmega: Sync to ArduPilot downstream 10-JAN-2022 (#1777)
27395aa7 2022-01-06 Julian Oes - pymavlink: update submodule to latest master (#1776)
2180611b 2022-01-06 Hamish Willee - MAV_WINCH_STATUS_CLUTCH_ENGAGED: Winch clutch is engaged allowing motor to move freely
7c6f104f 2022-01-05 Jacob Dahl - Extend WINCH_STATUS_FLAG and WINCH_ACTION (#1766)
dcb44e75 2022-01-06 Hamish Willee - Delete unused cmake tree (#1760)
5b8bec5b 2022-01-05 Hamish Willee - common.xml: AUTOTUNE_AXIS_YAW - fix value to correct one for bitmask (#1767)
2022-01-17 21:36:02 -05:00
Julian Oes be9385ef06 cmake: use at least 1 core to build SITL
If we use -j 0, ninja fails with invalid argument.
2022-01-17 17:07:31 -05:00
Daniel Agar 2fb82789fe lib/wind_estimator: pass q_att as const reference 2022-01-17 16:59:28 -05:00
Beat Küng 4cbee44220 rc.mc_defaults: only set IMU_GYRO_RATEMAX if default
Allows boards to set a higher rate.
2022-01-17 10:41:33 -05:00
Beat Küng 425b268feb boards: add holybro/kakuteh7
known issues:
- flash-based params does not work on H7 (due to ECC), so params are on
  SD card. The last flash sector is still reserved however.
- output channel 6 does not support DShot (the implementation does not
  handle channel gaps)
- flashing of the 2. flash bank is much slower (around 3x), than the 1.
  bank for some unknown reason.
- after the BL jumps to the app, there's several seconds passing until
  stm32_boardinitialize() is called.
2022-01-17 10:41:33 -05:00
Beat Küng cd2bb14f9b stm32h7/micro_hal.h: add RCC_APB1ENR_TIM4EN 2022-01-17 10:41:33 -05:00
Berker Canatar 9249d3ae3d Update current year in LICENSE 2022-01-17 08:10:57 +01:00
Silvan Fuhrer 0ce44cebb7 WindEstimator: use vehicle heading instead of ground course for initialisation
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2022-01-14 17:42:01 -05:00
Silvan Fuhrer 1a4b2b37bb AirspeedValidator: remove unused variable _tas_gate
Signed-off-by: Silvan Fuhrer <silvan@auterion.com>
2022-01-14 17:42:01 -05:00
Thomas Stastny 3b6c9448aa npfg: fix format CMakeLists 2022-01-14 22:05:53 +01:00
Thomas Stastny 13c36155ce fw_pos_ctrl: allow npfg to operate without wind estimate, dont switch controllers 2022-01-14 22:05:53 +01:00
Thomas Stastny 082e320191 integrate optional NPFG library for wind-aware fixed-wing guidance 2022-01-14 22:05:53 +01:00
Beat Küng 66cbbf9f2e control_allocator: use 'bidirectional' to clarify reversible motors 2022-01-14 12:18:29 -05:00
Beat Küng 9ca58f5e97 actuator outputs: always add reverse range param
and remove the possibility to set min > max to reverse.

Initially the idea was to add the checkbox on the UI side, to avoid adding
another param, but I don't think I'll go through the extra effort on the
QGC side.
2022-01-14 12:18:29 -05:00
Daniel Agar 98a9748bb8 drivers/dshot: use uORB::PublicationMulti for ORB_ID(esc_status)
- there can be multiple different outupt modules publishing esc_status (eg dshot & uavcan)
2022-01-14 12:17:49 -05:00
Daniel Agar 81ecd130fc uavcan: ESC publish esc_status in callback instead of timer
- this ensures every ESC status gets published with minimal latency
 - also prevents publishing ORB_ID(esc_status) when there's no actual
data (bug)
2022-01-14 12:17:49 -05:00
David Sidrane 25099ce11f fmurt1062-v1:Added GPS driver dropped in move to boardconfig 2022-01-14 07:04:06 -08:00
Julian Oes 651702c2c1 setup: don't install ccache
I don't think the PX4 setup script should decide whether a user is to
use ccache or not. Anecdotally, I've heard from some in the dev team
that they are not using it themselves, so I don't think there is a good
basis to push it onto others.

That being said, we can of course still use ccache as part of the CI
builds nevertheless.
2022-01-14 10:01:23 -05:00
Roman Bapst dc6067ed12 Update src/modules/fw_att_control/fw_att_control_params.c
Co-authored-by: Silvan Fuhrer <silvan@auterion.com>
2022-01-14 16:34:44 +03:00