mc_rate_control: use vehicle_angular_acceleration topic #14048

dagar · 2020-01-28T18:48:07Z

We're computing vehicle_angular_acceleration upstream of the controllers in the sensor pipeline now. Should we use it the multicopter rate controller?

allows us to see the data actually used by the controller
we'll be able to do a better job filtering data upstream
- if the mc_rate_controller isn't running synchronized with IMU sampling the upstream vehicle_angular_acceleration will still capture and filter all data
- in the future we can use the full raw data set (FIFOs) to filter as well
angular acceleration only goes through a single lowpass now, instead of applied after the lowpassed angular velocity
sample rate used by filter is now correct, especially if the primary gyro changes
saves some flash and cpu (minor)
downside, param churn: MC_DTERM_CUTOFF replaced by IMU_DGYRO_CUTOFF

jlecoeur

Sounds good to me.

Another benefit is that angular acceleration may be used for system ID (ex: #13585 ) , so it is better if the same data is used in the controller.

Requires careful flight test.

jlecoeur · 2020-01-30T18:06:07Z

src/modules/mc_rate_control/MulticopterRateControl.cpp

@@ -147,12 +145,18 @@ MulticopterRateControl::Run()
 	vehicle_angular_velocity_s angular_velocity;

 	if (_vehicle_angular_velocity_sub.update(&angular_velocity)) {
+
+		// grab corresponding vehicle_angular_acceleration immediately after vehicle_angular_velocity copy


We may document on both sides - _mc_rate_control and vehicle_angular_velocity - that velocity is published before accel, and that they should be read in the same order. This is not obvious and may be lost in a future refactoring.

The ordering doesn't actually matter because the execution of vehicle_angular_velocity (where both angular velocity and angular acceleration are published) followed by mc_rate_control is serialized. Although there's nothing actually enforcing this implementation.

I'll still add a comment publisher side.

Right, 👍

MaEtUgR

Makes sense from my pespective to handle this on the sensor side and I don't see the problem IF the filtering inside vehicle_angular_acceleration is equivalent which it should be.

Just came to my mind: What happens in the case where the gyro sensor is switched in air and the derivative should not have a jump?

Before we merge this we have to directly compare if it really doesn't change filtering because if there's a difference this could easily flip multicopters on takeoff and/or significantly impair performance.
Also @bkueng should be added as reviewer since he contributed the D-term filtering which was a big improvement.

MaEtUgR · 2020-02-02T07:22:22Z

src/modules/mc_rate_control/MulticopterRateControl.cpp

+
+		// grab corresponding vehicle_angular_acceleration immediately after vehicle_angular_velocity copy
+		vehicle_angular_acceleration_s v_angular_acceleration{};
+		_vehicle_angular_acceleration_sub.copy(&v_angular_acceleration);


Minor implementation detail but what about using a SubscriptionData and directly fetching the vector using:
const Vector3f angular_accl{_vhicle_angular_acceleration_sub.get().xyz}; ?

It's pretty minor, but there's no need to store the data because every iteration grabs a new one.

dagar · 2020-02-02T16:08:29Z

Just came to my mind: What happens in the case where the gyro sensor is switched in air and the derivative should not have a jump?

That's a good point. At the moment the filters aren't explicitly reset on sensor change, but may reset a few iterations later if the sample rate of the new primary is different (eg pixhawk 4 Invensense -> Bosch).

In most real world use the sensor failover is likely to be abrupt due to resetting the bias. This will be addressed once we have an estimator instance per IMU.

bkueng · 2020-02-03T07:59:18Z

I'm ok with the change, but please do:

set the default of IMU_DGYRO_CUTOFF to 30
improve param documentation:
- IMU_GYRO_CUTOFF is not applied to the d-term
- the notch is applied to both
- transfer current documentation of MC_DTERM_CUTOFF to IMU_DGYRO_CUTOFF (D-Term filter is a generally known term and should be kept).
make sure existing param values of MC_DTERM_CUTOFF are moved over to IMU_DGYRO_CUTOFF

…acceleration

…UTOFF)

Junkim3DR · 2020-02-04T20:33:14Z

Tested on NXP FMUK66 v3

Flight Card 1

Modes Tested
Position Mode: Good.
Altitude Mode: Good.
Stabilized Mode: Good.
Procedure
Arm and Take off in position mode, after flying for approximately one minute, switched to altitude then stabilized mode.
Notes
No issues noted, good flight in general.
Log
https://review.px4.io/plot_app?log=67181da9-5784-40f4-99cf-3d056e956be5

Flight Card 2

Modes Tested
Mission Plan Mode (Automated): Good.
RTL (Return To Land): Good.
Procedure
Arm and Take off in mission plan mode then make sure that vehicle follows all waypoints as shown in QGC, once all waypoints are completed make sure vehicle triggers RTL.
Notes
Good flight in general.
Log
https://review.px4.io/plot_app?log=6efd975c-1a9f-4a29-869f-37e3a69c9c85

Flight Card 3

Modes Tested
Position Mode: Good.
Mission Plan Mode (Automated): Good.
RTL (Return To Land): Good.
Procedure
Arm and Take off in position mode, after flying for approximately one minute, switched to mission plan mode then make sure that vehicle follows all waypoints as shown in QGC.
Notes
Good flight in general.
Log
https://review.px4.io/plot_app?log=afd62fec-6ca5-4e2e-99e4-a0d173616bae

jorge789 · 2020-02-04T21:11:58Z

Tested on CUAV nano V5

Flight Card 1

Modes Tested:
Position Mode: Good.
Altitude Mode: Good.
Stabilized Mode: Good.

Procedure:
Arm and Take off in position mode, after flying for approximately one minute, switched to altitude then stabilized mode.

Notes:
No issues were noted, good flight in general.

Logs: https://review.px4.io/plot_app?log=56fcea9c-9ef1-43dd-9356-33a19e8620d4

Flight Card 2

Modes Tested
Mission Plan Mode (Automated): Good.
RTL (Return To Land): Good.

Procedure

Armed form QGC.
Took-off on mission plan mode
. Note that commands were sent form QGC (Armed and takeoff)
Once all waypoint completed vehicle switched to RTL and landed as expected.
Good flight in general.
Note:
When the vehicle was landing, it did not descend vertically, moved away with the wind and returned to the takeoff point and landed

Logs: https://review.px4.io/plot_app?log=67168d92-dc09-44fb-b76c-eff8551accaa

Flight Card 3

Modes Tested
Position Mode: Good.
Mission Plan Mode (Automated): Good.
RTL (Return To Land): Good.

Procedure
Arm and Take off in position mode, after flying for approximately one minute, switched to mission plan mode then make sure that vehicle follows all waypoints as shown in QGC, once completed all waypoint then trigger RTL.

Good flight in general.

Note:
No issues were noted, good flight in general.

Logs: https://review.px4.io/plot_app?log=b6b3d5e0-e3c7-4b17-8558-7e9b257634cf

Flight Card 4

- Modes Tested
Position Mode: Good.
Failsafe RTL (Return To Land): Not returning after RC shutdown.
- Procedure
Arm and take off in position mode, after flying approximately a few seconds, turned off the RC to activate the fail-safe RTL, wait for the vehicle to return and land.
- Notes
No issue noted, good flight in general.

Logs: https://review.px4.io/plot_app?log=8b4b4dfb-3f95-4ead-99a0-22a2d13a982d

Tested on PixRacer V4

Flight Card 1

Modes Tested:
Position Mode: Good.
Altitude Mode: Good.
Stabilized Mode: Good.

Procedure:
Arm and Take off in position mode, after flying for approximately one minute, switched to altitude then stabilized mode.

Notes:
No issues were noted, good flight in general.

Logs: https://review.px4.io/plot_app?log=11f4eff7-7305-401c-8def-241c373316a9

Flight Card 2

Modes Tested
Mission Plan Mode (Automated): Good.
RTL (Return To Land): Good.

Procedure

Armed form QGC.
Took-off on mission plan mode
. Note that commands were sent form QGC (Armed and takeoff)
Once all waypoint completed vehicle switched to RTL and landed as expected.
Good flight in general.
Note:
No issues were noted, good flight in general.

Logs: https://review.px4.io/plot_app?log=668d6d5a-f34e-4e73-a489-7652b6c57936

Flight Card 3

Modes Tested
Position Mode: Good.
Mission Plan Mode (Automated): Good.
RTL (Return To Land): Good.

Procedure
Arm and Take off in position mode, after flying for approximately one minute, switched to mission plan mode then make sure that vehicle follows all waypoints as shown in QGC, once completed all waypoint then trigger RTL.
Note:
Good flight in general.

Note:
No issues were noted, good flight in general.

Logs: https://review.px4.io/plot_app?log=a5416a58-52eb-42e8-9d8a-7fce74716653

Flight Card 4

Modes Tested
Position Mode: Good.
Failsafe RTL (Return To Land): Not returning after RC shutdown.
Procedure
Arm and take off in position mode, after flying approximately a few seconds, turned off the RC to activate the fail-safe RTL, wait for the vehicle to return and land.
Notes
No issue noted, good flight in general.

Logs: https://review.px4.io/plot_app?log=e7f0e0ba-fcec-4ef0-960f-a5520e3ffb96

dannyfpv · 2020-02-04T21:35:48Z

Tested on pixhawk4 v5 f-450
Flight Card 1

Modes Tested
Position Mode: Good.
Altitude Mode: Good.
Stabilized Mode: Good.

Procedure
Arm and Take off in position mode, after flying for approximately one minute, switched to altitude then stabilized mode.

Notes
No issues noted, good flight in general.

Log
https://review.px4.io/plot_app?log=1b6c8ecf-11f5-40d9-9e5f-48424de475f7

Flight Card 2

Modes Tested
Mission Plan Mode (Automated): Good.
RTL (Return To Land): Good.

Procedure
Arm and Take off in mission plan mode then make sure that vehicle follows all waypoints as shown in QGC, once all waypoints are completed make sure vehicle triggers RTL.

Notes
Good flight in general.

Log
https://review.px4.io/plot_app?log=bcb3f7ab-dbec-4e3b-8f96-0137c5923038

Flight Card 3

Modes Tested
Position Mode: Good.
Mission Plan Mode (Automated): Good.
RTL (Return To Land): Good.

Procedure
Arm and Take off in position mode, after flying for approximately one minute, switched to mission plan mode then make sure that vehicle follows all waypoints as shown in QGC.

Notes
Good flight in general.

Log
https://review.px4.io/plot_app?log=5d23982a-bf11-4a5a-a464-304aaff9f24e

Flight Card 4

Modes Tested
Position Mode: Good.
Failsafe RTL (Return To Land): Not returning after RC shutdown.
Procedure
Arm and take off in position mode, after flying approximately a few seconds, turned off the RC to activate the fail-safe RTL, wait for the vehicle to return and land.
Notes
No issue noted, good flight in general.
Logs:
https://review.px4.io/plot_app?log=daa82187-c213-4a0e-a223-ec0a2e4955fc

…acceleration

dagar · 2020-02-06T01:38:59Z

Merge conflict resolved. @PX4/testflights could you do another multicopter flight with SDLOG_PROFILE "High rate" enabled. The new data we're talking about using here isn't logged by default. A single thorough multicopter flight on any vehicle with a comparison to master should suffice.

jorge789 · 2020-02-06T20:52:29Z

Tested on PixRacer V4

Flight Card 1

Modes Tested:
Position Mode: Good.
Altitude Mode: Good.
Stabilized Mode: Good.

Procedure:
Arm and Take off in position mode, after flying for approximately one minute, switched to altitude then stabilized mode.

Notes:
No issues were noted, good flight in general.

Logs: https://review.px4.io/plot_app?log=6c33763b-dd5f-4652-849b-d6591b9c74a0

Flight Card 2

Modes Tested
Mission Plan Mode (Automated): Good.
RTL (Return To Land): Good.

Procedure

Armed form QGC.
Took-off on mission plan mode
. Note that commands were sent form QGC (Armed and takeoff)
Once all waypoint completed vehicle switched to RTL and landed as expected.
Good flight in general.
Note:
No issues were noted, good flight in general.

Logs: https://review.px4.io/plot_app?log=50319699-7e95-4ae7-9b35-80f6bbded185

Flight Card 3

Modes Tested
Position Mode: Good.
Mission Plan Mode (Automated): Good.
RTL (Return To Land): Good.

Procedure
Arm and Take off in position mode, after flying for approximately one minute, switched to mission plan mode then make sure that vehicle follows all waypoints as shown in QGC, once completed all waypoint then trigger RTL.
Note:
Good flight in general.

Note:
No issues were noted, good flight in general.

Logs: https://review.px4.io/plot_app?log=d69037d1-b9ce-41e6-9755-6128d316c38a

Log
Master comparison
https://review.px4.io/plot_app?log=fa512e8c-2c4d-454e-9f45-5a71a2bf66fa

Junkim3DR · 2020-02-06T21:05:13Z

Tested on NXP FMUK66 v3

Flight Card 1

Modes Tested
Position Mode: Good.
Altitude Mode: Good.
Stabilized Mode: Good.
Procedure
Arm and Take off in position mode, after flying for approximately one minute, switched to altitude then stabilized mode.
Notes
No issues noted, good flight in general.
Log
https://review.px4.io/plot_app?log=bc935bc5-5e2f-4d6b-a644-d7323de12018

Flight Card 2

Modes Tested
Mission Plan Mode (Automated): Good.
RTL (Return To Land): Good.
ProcedureArm and Take off in mission plan mode then make sure that vehicle follows all waypoints as shown in QGC, once all waypoints are completed make sure vehicle triggers RTL.
Notes
Good flight in general.
Log
https://review.px4.io/plot_app?log=75dfd017-3691-48bc-baad-27551c7763c9

Flight Card 3

Modes Tested
Position Mode: Good.
Mission Plan Mode (Automated): Good.
RTL (Return To Land): Good.
Procedure
Arm and Take off in position mode, after flying for approximately one minute, switched to mission plan mode then make sure that vehicle follows all waypoints as shown in QGC.
Notes
Good flight in general.
Log
https://review.px4.io/plot_app?log=87d5e48a-9924-4e66-ac93-bd24a820bd94

Comparison Master

Modes Tested
Position Mode: Good.
Altitude Mode: Good.
Stabilized Mode: Good.
Mission Plan Mode (Automated): Good.
RTL (Return To Land): Good.
Procedure
Arm and Take off in position mode, after flying for approximately one minute, switched to altitude then stabilized mode proceed to switch to mission plan mode then make sure that vehicle follows all waypoints as shown in QGC, once completed all waypoint then trigger RTL.
Notes
No issues noted, good flight in general.
Log
https://review.px4.io/plot_app?log=4ae02f40-0287-49ee-a4ab-06d667bafdcb

…acceleration

dagar · 2020-03-12T14:04:40Z

Are we good to go here?

jlecoeur · 2020-03-12T17:10:12Z

Looks good to me. Le jeu. 12 mars 2020 à 15:04, Daniel Agar <[email protected]> a écrit :

…

Are we good to go here? — You are receiving this because your review was requested. Reply to this email directly, view it on GitHub <#14048 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAKQZB3FERDGFWYBLFC7M6TRHDTYRANCNFSM4KMXMIXA> .

MaEtUgR

I still consider it useful given the current filter producing the angular acceleration state is configured the same way as the derivative term filter before. Note that there was a sample rate detection (kind of related: #14303).

dagar · 2020-03-12T19:05:43Z

Thanks everyone. I'll get the documentation updated for the release.

bkueng · 2020-03-13T07:02:52Z

I'm missing the parameter transition support. It would make this considerably more user-friendly.

dagar · 2020-03-13T18:38:58Z

I'm missing the parameter transition support. It would make this considerably more user-friendly.

Good idea. #14389

dagar added Admin: Enhancement (improvement) 💡 Multirotor 🛸 Admin: Cleanup (housekeeping) 🧹 labels Jan 28, 2020

dagar requested review from jlecoeur and MaEtUgR January 28, 2020 18:48

dagar force-pushed the pr-mc_vehicle_angular_acceleration branch from ff83fcf to 1d7b01b Compare January 28, 2020 19:00

mc_rate_control: use vehicle_angular_acceleration topic

6357bd1

dagar force-pushed the pr-mc_vehicle_angular_acceleration branch from 699d1ea to 6357bd1 Compare January 30, 2020 07:09

jlecoeur previously approved these changes Jan 30, 2020

View reviewed changes

dagar requested a review from a team January 30, 2020 18:53

MaEtUgR reviewed Feb 2, 2020

View reviewed changes

MaEtUgR requested a review from bkueng February 2, 2020 07:32

weekly-digest bot mentioned this pull request Feb 2, 2020

Weekly Digest (26 January, 2020 - 2 February, 2020) #14084

Closed

dagar added 2 commits February 3, 2020 09:27

Merge remote-tracking branch 'px4/master' into pr-mc_vehicle_angular_…

601d464

…acceleration

IMU_DGYRO_CUTOFF change default to 30 Hz

f9e5d84

dagar dismissed jlecoeur’s stale review via f9e5d84 February 3, 2020 14:28

improve IMU_DGYRO_CUTOFF param documentation (updated from MC_DTERM_C…

4d106ae

…UTOFF)

dagar added 2 commits February 5, 2020 20:34

Merge remote-tracking branch 'px4/master' into pr-mc_vehicle_angular_…

4aa9630

…acceleration

sensors/vehicle_angular_velocity: angular acceleration LPF default 30 Hz

cab7757

weekly-digest bot mentioned this pull request Feb 9, 2020

Weekly Digest (2 February, 2020 - 9 February, 2020) #14124

Closed

dagar mentioned this pull request Mar 5, 2020

SITL jmavsim crazy pitch rate oscillations #14303

Closed

dagar added this to the Release v1.11.0 milestone Mar 5, 2020

dagar added 3 commits March 9, 2020 08:20

Merge remote-tracking branch 'px4/master' into pr-mc_vehicle_angular_…

5db28ff

…acceleration

Merge branch 'master' into pr-mc_vehicle_angular_acceleration

96e15c8

Merge branch 'master' into pr-mc_vehicle_angular_acceleration

67cea7c

MaEtUgR approved these changes Mar 12, 2020

View reviewed changes

dagar merged commit 093e9ba into master Mar 12, 2020

dagar deleted the pr-mc_vehicle_angular_acceleration branch March 12, 2020 19:07

dagar mentioned this pull request Mar 13, 2020

parameters: handle MC_DTERM_CUTOFF -> IMU_DGYRO_CUTOFF migration #14389

Merged

dagar mentioned this pull request Mar 31, 2020

Racer Setup: Fix removal of MC_DTERM_CUTOFF PX4/PX4-user_guide#703

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

mc_rate_control: use vehicle_angular_acceleration topic #14048

mc_rate_control: use vehicle_angular_acceleration topic #14048

dagar commented Jan 28, 2020 •

edited

Loading

jlecoeur left a comment •

edited

Loading

jlecoeur Jan 30, 2020

dagar Jan 30, 2020

jlecoeur Jan 30, 2020

MaEtUgR left a comment •

edited

Loading

MaEtUgR Feb 2, 2020

dagar Feb 2, 2020

dagar commented Feb 2, 2020

bkueng commented Feb 3, 2020

Junkim3DR commented Feb 4, 2020

jorge789 commented Feb 4, 2020

dannyfpv commented Feb 4, 2020

dagar commented Feb 6, 2020

jorge789 commented Feb 6, 2020

Junkim3DR commented Feb 6, 2020

dagar commented Mar 12, 2020

jlecoeur commented Mar 12, 2020 via email

MaEtUgR left a comment •

edited

Loading

dagar commented Mar 12, 2020

bkueng commented Mar 13, 2020

dagar commented Mar 13, 2020

mc_rate_control: use vehicle_angular_acceleration topic #14048

mc_rate_control: use vehicle_angular_acceleration topic #14048

Conversation

dagar commented Jan 28, 2020 • edited Loading

jlecoeur left a comment • edited Loading

Choose a reason for hiding this comment

jlecoeur Jan 30, 2020

Choose a reason for hiding this comment

dagar Jan 30, 2020

Choose a reason for hiding this comment

jlecoeur Jan 30, 2020

Choose a reason for hiding this comment

MaEtUgR left a comment • edited Loading

Choose a reason for hiding this comment

MaEtUgR Feb 2, 2020

Choose a reason for hiding this comment

dagar Feb 2, 2020

Choose a reason for hiding this comment

dagar commented Feb 2, 2020

bkueng commented Feb 3, 2020

Junkim3DR commented Feb 4, 2020

Tested on NXP FMUK66 v3

jorge789 commented Feb 4, 2020

Tested on CUAV nano V5

Tested on PixRacer V4

dannyfpv commented Feb 4, 2020

dagar commented Feb 6, 2020

jorge789 commented Feb 6, 2020

Tested on PixRacer V4

Junkim3DR commented Feb 6, 2020

Tested on NXP FMUK66 v3

dagar commented Mar 12, 2020

jlecoeur commented Mar 12, 2020 via email

MaEtUgR left a comment • edited Loading

Choose a reason for hiding this comment

dagar commented Mar 12, 2020

bkueng commented Mar 13, 2020

dagar commented Mar 13, 2020

dagar commented Jan 28, 2020 •

edited

Loading

jlecoeur left a comment •

edited

Loading

MaEtUgR left a comment •

edited

Loading

MaEtUgR left a comment •

edited

Loading