PX4 Dev Call: July 20, 2022

rroche · July 13, 2022, 5:00pm

July 20, 2022

Join us

Agenda

Community Announcement
Community Q&A
Project Updates
General Discussions
Weekly Overview
High priority queue
Release

Community Announcement

PX4 Dev Summit video

Will be premiered one at a time ! First video will be the Keynote

Community Q&A (No deep technical discussions)

Guideline for asking a Question

Specify what you are trying to achieve
Specify what environment / platform you are using
Prepare a Link or Document to help understand if possible
First, ask the question on the Slack or create a Github Issue!
If you take over 5 minutes for the question, please continue in Slack or a Github Issue.

Q.1 : Testing lightware LiDaR - Andrew B

Need helping testing lightware LiDAR with collision prevention, need power module (I already have one) and flight controller with at least 2 telem ports (my second one can’t transmit out). Daniel is helping

github.com/PX4/PX4-Autopilot

LightwareLidar SF45/B rotating sensor serial driver

PX4:main ← dirksavage88:lightware_sf45

opened 03:42PM - 08 Jul 22 UTC

dirksavage88

+1256 -0

Lightware SF45/B rotating distance sensor driver that publishes distance and yaw… angle (through the obstacle distance sensor message). Preliminary testing I show the sensor is able to detect obstacles: ![Screenshot from 2022-12-29 11-52-46](https://user-images.githubusercontent.com/35986980/210007286-257af903-7d7e-4332-865a-606053440c08.png) ![Screenshot from 2022-07-08 10-14-28](https://user-images.githubusercontent.com/35986980/178021450-a02bed6b-8676-461c-8670-b361adf25a0f.png) ![Screenshot from 2022-07-08 10-14-35](https://user-images.githubusercontent.com/35986980/178021479-f3f10f04-d022-4cd5-ba83-3c5181ba76c0.png) I am able to publish to obstacle distance msg and I see valid distance values moving through the bins. It is also stopping before collision with an obstacle if within CP distance defined in the parameters (tested using 3m). Ardupilot has a working driver that can be used as a reference [here](https://github.com/ArduPilot/ardupilot/blob/master/libraries/AP_Proximity/AP_Proximity_LightWareSF45B.cpp) Additionally, Lightware has a sample cpp library on their github. Serial driver defaults to TEL1 due to the power requirements of the sensor which is documented [here](http://support.lightware.co.za/sf45b/#/pixhawk) It is possible to change this default for newer boards that have other uarts that satisfy the power requirements, this can be done through SENS_EN_SF45 parameter. Baud rate is set to 115200. Update rate defaults to 50hz, but I have tested at higher rates. This can be set by parameter. Orientation (upright or upside down mounting) parameter is available, will have to verify the yaw corrections for the upside down mounting is correct. There is also a yaw configuration parameter that allows user to have sensor face in a different direction other than default forward facing. An I2C driver can be implemented but does not permit continuous streaming of distance data. The lightware sf45 can be configured in the Lightware Studio [software](https://lightwarelidar.com/pages/lightware-studio) to set the Baud rate before connecting to Pixhawk, additionally the scan angle can be changed for the largest field of view possible: 0-160 and -160 to 0 ![Screenshot from 2022-07-08 10-32-15](https://user-images.githubusercontent.com/35986980/178024065-f837572a-50d4-44d6-a7b5-ab49f6b3d71f.png) Future considerations: Parameters to map sectors around the airframe to ignore (e.g. gps mast). @dagar I can send you the sensor if you would like your own to test. @davids5 the serial port local options appear to work, let me know if it needs refinement or extra comments to explain the serial options configuration

Q.2 : Custom Multicopter control using dynamic control allocation - Salvatore

My question is about the control allocation for custom drones. In my research lab we have different types of multirotor, for example i’m working on a tilting rotor coaxial octocopter and i need to allocate both the speed of the propeller and the angles of the servomotors.

This is something that is not already present in the firmware (if i’m not wrong).
Is there a guide (or will be) on how to write a control allocator/effectivness matrix? All of our drones works with px4 but we have this little-big limitation

Right now working in X-shaped, Tilt-rotor coaxial quadrotor

Response from Mathieu on Slack:
You can simply take as example one of the existing ones (Pseudo-Inverse, sequential desaturation) and add your own algorithm here:

And then add it here:

github.com

PX4/PX4-Autopilot/blob/2f3cb97872fdfdede519f0ce4bfca35e9f4e16ab/src/modules/control_allocator/ControlAllocator.cpp#L179-L191


      
          			switch (method) {
          			case AllocationMethod::PSEUDO_INVERSE:
          				_control_allocation[i] = new ControlAllocationPseudoInverse();
          				break;
          
          			case AllocationMethod::SEQUENTIAL_DESATURATION:
          				_control_allocation[i] = new ControlAllocationSequentialDesaturation();
          				break;
          
          			default:
          				PX4_ERR("Unknown allocation method");
          				break;
          			}

Dagar: Get it to a level that’s functional (Simulation / ), and ping him on slack

Project Updates

P.1 : Replace fixedwing ECL rate controller with RateControl Library #19912 - Jay

github.com/PX4/PX4-Autopilot

Replace fixedwing ECL rate controller with RateControl Library

PX4:main ← PX4:pr-fw-rate-control

opened 12:16PM - 13 Jul 22 UTC

Jaeyoung-Lim

+309 -400

## Describe problem solved by this pull request Yaw rate was previously not con…trolled for fixed wings. This made it hard to integrate offboard control methods that might want to control body rates (and eventually yaw rates) directly. This is due to the fact that fixed wing attitude control uses separate ECL controllers for each axis (roll, pitch, yaw) and specifically yaw rate control is only calculated using the current roll to maintain coordinated flight. ## Describe your solution This commit makes the `fw_att_control` share the rate controller as a library with the `mc_rate_control` module. - Enables passing yaw rate setpoints directly in offboard mode - Allows controlling yaw rate setpoints in acro mode through the sticks - Allows controlling yaw rate setpoints in offboard mode - Removes euler body rate control from the ECL controllers - By using the rate control library, the fw attitude controller now uses angular acceleration estimates for rate control The attitude control and turn coordination is still done through ECL controllers - [x] Add airspeed scaling - [x] Improve tracking performance (Done by @sfuhrer ) ## Describe possible alternatives A clear and concise description of alternative solutions or features you've considered. ## Test data / coverage - Tested in SITL Gazebo: [flight log](https://review.px4.io/plot_app?log=9a7a4bf1-fe63-4766-b1c5-250086de104a) ``` make px4_sitl gazebo_plane ``` - NOT tested on hardware ## Additional context fixes https://github.com/PX4/PX4-Autopilot/issues/12886

Discussion on Airspeed’s effect on the control
Silvan : What is preventing us from having a unified rate controller (for Fixed Wing, Multirotor, etc)? : Airspeed
Now the rate controller doesn’t do the turn coordination
Currently Multicopter Rate Controller has a hacky I-term modification
Would be better to have a unified rate controller that can abstract-ify the vehicle type.
Current VTOL using both Fixed-Wing & Multicopter rate controller created hard-switching between the controllers, so having a unified controller would be great.

Flight Testings - Farhang

Tests are coming this week. Got stuck on QGC GPS disabled, after finding out the issue I will continue and share. Just an update

P.2 : PX4 Failsafe overhaul RFC from Kaylan

Please review if you have time!
Also, Kalyan is interested in doing a broad pass no DroneCAN in PX4, so if you have thoughts / ideas, contact him!

General Discussions

D.1 : Bringing back the pwm test functionality

There was a Slack Thread on PX4 now missing the ability to do pwm test commands to send raw signals (e.g. 2000 us PWM value to a motor) via CLI.

There is an actuator_test command, but this only works if the user has mapped the functions to the outputs (e.g. Motor1, Servo1, etc), which is commonly not the case for low level hardware developers who wants to quickly test out the PWM output without setting all of these high level concepts.

Ideas

Do the #ifdef to only include the pwm test when doing hardware tests
Create a ‘timer test’

There’s a branch that aims to implement the pwm_test functionality back

D.2 : Discuss Forum not being active and Slack’s history getting deleted problem

Currently the discuss forum is quite inactive and a lot of users don’t get feedback. Whereas Slack is more active but we are on a free plan so the history is only preserved up to 10,000 recent messages (which is around 1~2 month for the size of PX4 Slack community.

Using Discuss Forum, we have the whole knowledge preserved, but since it isn’t as convenient as the Slack, not a lot of users get fast feedback.

And Slack is great for having fast feedbacks but all the conversation gets deleted after 1~2 months, which is terrible in terms of preserving the knowledge & benefits of future users.

Migration plan of the messaging platform is probably not ideal, since it’s worse than Slack.

Solution:

Create a Slack Bot that can nudge people to move discussion to Discuss Forum when the thread gets long
Have a Bot that can migrate the thread of conversations automatically to Discuss Forum when the user manually presses the button.

D.3 : Reducing flash usage - David

The 6X PR is having flash overflowing

Reducing the number of airframes as we now have the control allocation framework
Right now in VTOL, Flaperons aren’t working without the specific airframe definition (/mixer)
Currently a lot of the airframes aren’t very helpful (very custom configurations)

Weekly Overview

Github

Recently updated Issues / PRs in PX4-Autopilot Link

Pull Requests

Replace fixedwing ECL rate controller with RateControl Library by Jaeyoung-Lim · Pull Request #19912 · PX4/PX4-Autopilot · GitHub
Issues