docs(updates): Link fixes and orphan image removals (#26925)

* docs(updates): Link fixes and orphan image removals
* docs(delete): Remove orphaned translations
This commit is contained in:
Hamish Willee
2026-04-01 13:51:06 +11:00
committed by GitHub
parent 5568c66959
commit 4cbdc3dec6
40 changed files with 118 additions and 1262 deletions
-86
View File
@@ -1,86 +0,0 @@
# Neural Network Module: System Integration
The neural control module ([mc_nn_control](../modules/modules_controller.md#mc-nn-control)) implements an end-to-end controller utilizing neural networks.
The parts of the module directly concerned with generating the code for the trained neural network and integrating it into the module are covered in [TensorFlow Lite Micro (TFLM)](../advanced/tflm.md).
This page covers the changes that were made to integrate the module into PX4, both within the module, and in larger system configuration.
:::tip
This topic should help you to shape the module to your own needs.
You will need some familiarity with PX4 development.
For more information see the developer [Getting Started](../dev_setup/getting_started.md).
:::
## 자동 실행
A line to autostart the [mc_nn_control](../modules/modules_controller.md#mc-nn-control) module has been added in the [`ROMFS/px4fmu_common/init.d/rc.mc_apps`](https://github.com/PX4/PX4-Autopilot/blob/main/ROMFS/px4fmu_common/init.d/rc.mc_apps) startup script.
It checks whether the module is included by looking for the parameter [MC_NN_EN](../advanced_config/parameter_reference.md#MC_NN_EN).
If this is set to `1` (the default value), the module will be started when booting PX4.
Similarly you could create other parameters in the [`mc_nn_control_params.c`](https://github.com/PX4/PX4-Autopilot/blob/main/src/modules/mc_nn_control/mc_nn_control_params.c) file for other startup script checks.
## Custom Flight Mode
The module creates its own flight mode "Neural Control" which lets you choose it from the flight mode menu in QGC and bind it to a switch on you RC controller.
This is done by using the [ROS 2 Interface Library](../ros2/px4_ros2_interface_lib.md) internally.
This involves several steps and is visualized here:
:::info
The module does not actually use ROS 2, it just uses the API exposed through uORB topics.
:::
:::info
In some QGC versions the flight mode does not show up, so make sure to update to the newest version.
This only works for some flight controllers, so you might have to use an RC controller to switch to the correct external flight mode.
:::
![neural_mode_registration](../../assets/advanced/neural_mode_registration.png)
1. Publish a [RegisterExtComponentRequest](../msg_docs/RegisterExtComponentRequest.md).
This specifies what you want to create, you can read more about this in the [Control Interface](../ros2/px4_ros2_control_interface.md).
In this case we register an arming check and a mode.
2. Wait for a [RegisterExtComponentReply](../msg_docs/RegisterExtComponentReply.md).
This will give feedback on wether the mode registration was successful, and what the mode and arming check id is for the new mode.
3. [Optional] With the mode id, publish a [VehicleControlMode](../msg_docs/VehicleControlMode.md) message on the `config_control_setpoints` topic.
Here you can configure what other modules run in parallel.
The example controller replaces everything, so it turns off allocation.
If you want to replace other parts you can enable or disable the modules accordingly.
4. [Optional] With the mode id, publish a [ConfigOverrides](../msg_docs/ConfigOverrides.md) on the `config_overrides_request` topic.
(This is not done in the example module) This will let you defer failsafes or stop it from automatically disarming.
5. When the mode has been registered a [ArmingCheckRequest](../msg_docs/ArmingCheckRequest.md) will be sent, asking if your mode has everything it needs to run.
This message must be answered with a [ArmingCheckReply](../msg_docs/ArmingCheckReply.md) so the mode is not flagged as unresponsive.
In this response it is possible to set what requirements the mode needs to run, like local position.
If any of these requirements are set the commander will stop you from switching to the mode if they are not fulfilled.
It is also important to set health_component_index and num_events to 0 to not get a segmentation fault.
Unless you have a health component or events.
6. Listen to the [VehicleStatus](../msg_docs/VehicleStatus.md) topic.
If the nav_state equals the assigned `mode_id`, then the Neural Controller is activated.
7. When active the module will run the controller and publish to [ActuatorMotors](../msg_docs/ActuatorMotors.md).
If you want to replace a different part of the controller, you should find the appropriate topic to publish to.
To see how the requests are handled you can check out [src/modules/commander/ModeManagement.cpp](https://github.com/PX4/PX4-Autopilot/blob/main/src/modules/commander/ModeManagement.cpp).
## 로깅
To add module-specific logging a new topic has been added to [uORB](../middleware/uorb.md) called [NeuralControl](../msg_docs/NeuralControl.md).
The message definition is also added in `msg/CMakeLists.txt`, and to [`src/modules/logger/logged_topics.cpp`](https://github.com/PX4/PX4-Autopilot/blob/main/src/modules/logger/logged_topics.cpp) under the debug category.
For these messages to be saved in your logs you need to include `debug` in the [SDLOG_PROFILE](../advanced_config/parameter_reference.md#SDLOG_PROFILE) parameter.
## Timing
The module has two includes for measuring the inference times.
The first one is a driver that works on the actual flight controller units, but a second one, `chrono`, is loaded for SITL testing.
Which timing library is included and used is based on wether PX4 is built with NUTTX or not.
## Changing the setpoint
The module uses the [TrajectorySetpoint](../msg_docs/TrajectorySetpoint.md) message’s position fields to define its target.
To follow a trajectory, you can send updated setpoints.
For an example of how to do this in a PX4 module, see the [mc_nn_testing](https://github.com/SindreMHegre/PX4-Autopilot-public/tree/main/src/modules/mc_nn_testing) module in this fork.
Note that this is not included in upstream PX4.
To use it, copy the module folder from the linked repository into your workspace, and enable it by adding the following line to your `.px4board` file:
```sh
CONFIG_MODULES_MC_NN_TESTING=y
```
-77
View File
@@ -1,77 +0,0 @@
# TensorFlow Lite Micro (TFLM)
The PX4 [Multicopter Neural Network](../advanced/neural_networks.md) module ([mc_nn_control](../modules/modules_controller.md#mc-nn-control)) integrates a neural network that uses the [TensorFlow Lite Micro (TFLM)](https://github.com/tensorflow/tflite-micro) inference library.
This is a mature inference library intended for use on embedded devices, and is hence a suitable choice for PX4.
This guide explains how the TFLM library is integrated into the [mc_nn_control](../modules/modules_controller.md#mc-nn-control) module, and the changes you would have to make to use it for your own neural network.
:::tip
For more information, see the [TFLM guide](https://ai.google.dev/edge/litert/microcontrollers/get_started).
:::
## TLMF NN Formats
TFLM uses networks in its own [tflite format](https://ai.google.dev/edge/litert/models/convert).
However, since many microcontrollers do not have native filesystem support, a tflite file can be converted to a C++ source and header file.
This is what is done in `mc_nn_control`.
The tflight neural network is represented in code by the files [`control_net.cpp`](https://github.com/PX4/PX4-Autopilot/blob/main/src/modules/mc_nn_control/control_net.cpp) and [`control_net.hpp`](https://github.com/PX4/PX4-Autopilot/blob/main/src/modules/mc_nn_control/control_net.hpp).
### Getting a Network in tflite Format
There are many online resource for generating networks in the `.tflite` format.
For this example we trained the network in the open source [Aerial Gym Simulator](https://ntnu-arl.github.io/aerial_gym_simulator/).
Aerial Gym includes a guide, and supports RL both for control and vision-based navigation tasks.
The project includes conversion code for `PyTorch -> TFLM` in the [resources/conversion](https://github.com/ntnu-arl/aerial_gym_simulator/tree/main/resources/conversion) folder.
### Updating `mc_nn_control` with your own NN
You can convert a `.tflite` network into a `.cc` file in the ubuntu terminal with this command:
```sh
xxd -i converted_model.tflite > model_data.cc
```
You will then have to modify the `control_net.hpp` and `control_net.cpp` to include the data from `model_data.cc`:
- Take the size of the network in the bottom of the `.cc` file and replace the size in `control_net.hpp`.
- Take the data in the model array in the `cc` file, and replace the ones in `control_net.cpp`.
You are now ready to run your own network.
## Code Explanation
This section explains the code used to integrate the NN in `control_net.cpp`.
### Operations and Resolver
Firstly we need to create the resolver and load the needed operators to run inference on the NN.
This is done in the top of `mc_nn_control.cpp`.
The number in `MicroMutableOpResolver<3>` represents how many operations you need to run the inference.
A full list of the operators can be found in the [micro_mutable_op_resolver.h](https://github.com/tensorflow/tflite-micro/blob/main/tensorflow/lite/micro/micro_mutable_op_resolver.h) file.
There are quite a few supported operators, but you will not find the most advanced ones.
In the control example the network is fully connected so we use `AddFullyConnected()`.
Then the activation function is ReLU, and we `AddAdd()` for the bias on each neuron.
### Interpreter
In the `InitializeNetwork()` we start by setting up the model that we loaded from the source and header file.
Next is to set up the interpreter, this code is taken from the TFLM documentation and is thoroughly explained there.
The end state is that the `_control_interpreter` is set up to later run inference with the `Invoke()` member function.
The `_input_tensor` is also defined, it is fetched from `_control_interpreter->input(0)`.
### 입력
The `_input_tensor` is filled in the `PopulateInputTensor()` function.
`_input_tensor` works by accessing the `->data.f` member array and fill in the required inputs for your network.
The inputs used in the control network is covered in [Neural Networks](../advanced/neural_networks.md).
### 출력
For the outputs the approach is fairly similar to the inputs.
After setting the correct inputs, calling the `Invoke()` function the outputs can be found by getting `_control_interpreter->output(0)`.
And from the output tensor you get the `->data.f` array.
@@ -1,162 +0,0 @@
# _Pixhawk 4 Mini_ Wiring Quick Start
:::warning
PX4 does not manufacture this (or any) autopilot.
Contact the [manufacturer](https://holybro.com/) for hardware support or compliance issues.
:::
This quick start guide shows how to power the [_Pixhawk<sup>&reg;</sup> 4 Mini_](../flight_controller/pixhawk4_mini.md) flight controller and connect its most important peripherals.
![Pixhawk4 mini](../../assets/flight_controller/pixhawk4mini/pixhawk4mini_iso_1.png)
## 배선 개요
아래의 이미지는 주요 센서와 주변 장치(모터 및 서보 출력 제외)의 연결 방법을 설명합니다.
![Pixhawk 4 Mini Wiring Overview](../../assets/flight_controller/pixhawk4mini/pixhawk4mini_wiring_overview.png)
:::tip
More information about available ports can be found here: [_Pixhawk 4 Mini_ > Interfaces](../flight_controller/pixhawk4_mini.md#interfaces).
:::
## 콘트롤러 장착 및 장착 방향
_Pixhawk 4 Mini_ should be mounted on your frame using vibration-damping foam pads (included in the kit).
차량의 무게 중심에 최대한 가깝운 프레임에 장착하여야 하며, 화살표가 차량의 앞쪽과 위쪽을 향하도록 하여야 합니다.
![Pixhawk 4 Mini Orientation](../../assets/flight_controller/pixhawk4mini/pixhawk4mini_orientation.png)
:::info
If the controller cannot be mounted in the recommended/default orientation (e.g. due to space constraints) you will need to configure the autopilot software with the orientation that you actually used: [Flight Controller Orientation](../config/flight_controller_orientation.md).
:::
## GPS + 나침반 + 부저 + 안전 스위치 + LED
Attach the provided GPS with integrated compass, safety switch, buzzer, and LED to the **GPS MODULE** port. The GPS/Compass should be [mounted on the frame](../assembly/mount_gps_compass.md) as far away from other electronics as possible, with the direction marker towards the front of the vehicle (separating the compass from other electronics will reduce interference).
![Connect compass/GPS to Pixhawk 4](../../assets/flight_controller/pixhawk4mini/pixhawk4mini_gps.png)
:::info
The GPS module's integrated safety switch is enabled _by default_ (when enabled, PX4 will not let you arm the vehicle).
비활성화하려면 안전 스위치를 1초간 길게 누르십시오.
안전 스위치를 다시 눌러 안전 장치를 활성화하고 기체 시동을 끌 수 있습니다.
조종기나 지상국 프로그램에서 기체 시동을 끌 수 없는 상황에서 유용합니다.
:::
## 전원
PMB(Power Management Board)는 배전 보드와 전원 모듈로 사용됩니다.
In addition to providing regulated power to _Pixhawk 4 Mini_ and the ESCs, it sends information to the autopilot about the battery’s voltage and current draw.
Connect the output of the PMB that comes with the kit to the **POWER** port of the _Pixhawk 4 Mini_ using a 6-wire cable.
ESC와 서보에 대한 전원 공급 및 신호 연결을 위한 전원관리보드의 연결 방법은 아래의 표에서 설명합니다.
![Pixhawk 4 - Power Management Board](../../assets/flight_controller/pixhawk4mini/pixhawk4mini_power_management.png)
:::info
The image above only shows the connection of a single ESC and a single servo.
나머지 ESC와 서보를 비슷하게 연결합니다.
:::
| 핀 또는 커넥터 | 기능 |
| -------- | -------------------------------------------------------------------------- |
| B+ | ESC에 전원을 공급하기 위해 ESC B +에 연결 |
| GND | ESC 접지에 연결 |
| PWR | JST-GH 6-pin Connector, 5V 3A output<br> connect to _Pixhawk 4 Mini_ POWER |
| BAT | 전원 입력, 2 ~ 12S LiPo 배터리에 연결 |
The pinout of the _Pixhawk 4 Mini_ **POWER** port is shown below.
The `CURRENT` signal should carry an analog voltage from 0-3.3V for 0-120A as default.
The `VOLTAGE` signal should carry an analog voltage from 0-3.3V for 0-60V as default.
VCC 라인은 최소 3A 연속을 제공하여야하며, 기본적으로 5.1V로 설정되어야 합니다. 5V 보다 낮은 전압은 권장되지 않습니다.
| 핀 | 신호 | 전압 |
| --------------------------- | ------- | --------------------- |
| 1(red) | VCC | +5V |
| 2(black) | VCC | +5V |
| 3(black) | CURRENT | +3.3V |
| 4(black) | VOLTAGE | +3.3V |
| 5(black) | GND | GND |
| 6(black) | GND | GND |
:::info
If using a plane or rover, the 8 pin power (+) rail of **MAIN OUT** will need to be separately powered in order to drive servos for rudders, elevons, etc.
전원 레일을 BEC가 장착된 ESC 또는 독립형 5V BEC 또는 2S LiPo 배터리에 연결하여야 합니다.
서보에 제공되는 전압이 적절한 지 체크하십시오.
:::
<!--In the future, when Pixhawk 4 kit is available, add wiring images/videos for different airframes.-->
:::info
Using the Power Module that comes with the kit you will need to configure the _Number of Cells_ in the [Power Settings](https://docs.qgroundcontrol.com/master/en/qgc-user-guide/setup_view/power.html) but you won't need to calibrate the _voltage divider_.
You will have to update the _voltage divider_ if you are using any other power module (e.g. the one from the Pixracer).
:::
## 무선 조종
A remote control (RC) radio system is required if you want to _manually_ control your vehicle (PX4 does not require a radio system for autonomous flight modes).
You will need to [select a compatible transmitter/receiver](../getting_started/rc_transmitter_receiver.md) and then _bind_ them so that they communicate (read the instructions that come with your specific transmitter/receiver).
The instructions below show how to connect the different types of receivers to _Pixhawk 4 Mini_:
- Spektrum/DSM or S.BUS receivers connect to the **DSM/SBUS RC** input.
![Pixhawk 4 Mini - Radio port for Spektrum receivers](../../assets/flight_controller/pixhawk4mini/pixhawk4mini_rc_dsmsbus.png)
- PPM receivers connect to the **PPM RC** input port.
![Pixhawk 4 Mini - Radio port for PPM receivers](../../assets/flight_controller/pixhawk4mini/pixhawk4mini_rc_ppm.png)
- PPM and PWM receivers that have an _individual wire for each channel_ must connect to the **PPM RC** port _via a PPM encoder_ [like this one](https://www.getfpv.com/radios/radio-accessories/holybro-ppm-encoder-module.html) (PPM-Sum receivers use a single signal wire for all channels).
For more information about selecting a radio system, receiver compatibility, and binding your transmitter/receiver pair, see: [Remote Control Transmitters & Receivers](../getting_started/rc_transmitter_receiver.md).
## Telemetry Radio (Optional)
무선 텔레메트리는 지상국 프로그램에서 비행 차량의 통신/제어에 사용합니다(예 : UAV를 특정 위치로 지시하거나 새 임무를 업로드 할 수 있음).
The vehicle-based radio should be connected to the **TELEM1** port as shown below (if connected to this port, no further configuration is required).
다른 텔레메트리는 일반적으로 지상국 컴퓨터나 모바일 장치에 USB를 통하여 연결됩니다.
![Pixhawk 4 Mini Telemetry](../../assets/flight_controller/pixhawk4mini/pixhawk4mini_telemetry.png)
## micro SD 카드 (선택 사항)
SD cards are highly recommended as they are needed to [log and analyse flight details](../getting_started/flight_reporting.md), to run missions, and to use UAVCAN-bus hardware.
Insert the card (included in the kit) into _Pixhawk 4 Mini_ as shown below.
![Pixhawk 4 Mini SD Card](../../assets/flight_controller/pixhawk4mini/pixhawk4mini_sdcard.png)
:::tip
For more information see [Basic Concepts > SD Cards (Removable Memory)](../getting_started/px4_basic_concepts.md#sd-cards-removable-memory).
:::
## 모터
Motors/servos are connected to the **MAIN OUT** ports in the order specified for your vehicle in the [Airframe Reference](../airframes/airframe_reference.md). See [_Pixhawk 4 Mini_ > Supported Platforms](../flight_controller/pixhawk4_mini.md#supported-platforms) for more information.
:::info
이 참고사항은 모든 지원되는 기체 프레임의 출력 포트의 모터/서보 연결 리스트입니다. 프레임이 참고사항에 기재되어 있지 않다면, 올바른 유형의 "일반" 프레임을 사용하십시오.
:::
:::warning
The mapping is not consistent across frames (e.g. you can't rely on the throttle being on the same output for all plane frames).
가지고 있는 기체의 프레임에 대해 올바르게 모터를 제대로 연결하였는지 다시 한 번 확인하십시오.
:::
## 기타 주변 장치
The wiring and configuration of optional/less common components is covered within the topics for individual [peripherals](../peripherals/index.md).
## 설정
General configuration information is covered in: [Autopilot Configuration](../config/index.md).
QuadPlane specific configuration is covered here: [QuadPlane VTOL Configuration](../config_vtol/vtol_quad_configuration.md)
<!-- Nice to have detailed wiring infographic and instructions for different vehicle types. -->
## 추가 정보
- [_Pixhawk 4 Mini_](../flight_controller/pixhawk4_mini.md)
-7
View File
@@ -1,7 +0,0 @@
# DroPix Flight Controller (Discontinued)
<Badge type="info" text="Discontinued" />
The Drotek<sup>&reg;</sup> _DroPix autopilot_ is no longer available on the Drotek website, and is assumed to be discontinued.
See [PX4 v1.13 Documentation > DroPix Flight Controller](https://docs.px4.io/v1.13/en/flight_controller/dropix.html) for documentation.
-52
View File
@@ -1,52 +0,0 @@
# Test MC_07 - VIO (Visual-Inertial Odometry)
## Objective
To test that external vision (VIO) works as expected
## Preflight
Disconnect all GPS / compasses and ensure vehicle is using VIO for navigation
Ensure that the drone can go into Altitude / Position flight mode while still on the ground
Ensure there are no other sources of positioning besides VIO:
- [EKF2_OF_CTRL](../advanced_config/parameter_reference.md#EKF2_OF_CTRL): `0`
- [EKF2_GPS_CTRL](../advanced_config/parameter_reference.md#EKF2_GPS_CTRL): `0`
- [EKF2_EV_CTRL](../advanced_config/parameter_reference.md#EKF2_EV_CTRL): `15`
- [SYS_HAS_MAG](../advanced_config/parameter_reference.md#SYS_HAS_MAG): `0`
## Flight Tests
❏ Altitude flight mode
&nbsp;&nbsp;&nbsp;&nbsp;❏ Vertical position should hold current value with stick centered
&nbsp;&nbsp;&nbsp;&nbsp;❏ Pitch/Roll/Yaw response 1:1
&nbsp;&nbsp;&nbsp;&nbsp;❏ Throttle response set to climb/descent rate
❏ Position flight mode
&nbsp;&nbsp;&nbsp;&nbsp;❏ Horizontal position should hold current value with stick centered
&nbsp;&nbsp;&nbsp;&nbsp;❏ Vertical position should hold current value with stick centered
&nbsp;&nbsp;&nbsp;&nbsp;❏ Throttle response set to climb/descent rate
&nbsp;&nbsp;&nbsp;&nbsp;❏ Pitch/Roll/Yaw response set to pitch/roll/yaw rates
## 착륙
❏ Land in either Position or Altitude mode with the throttle below 40%
❏ Upon touching ground, copter should disarm automatically within 2 seconds (default: see [COM_DISARM_LAND](../advanced_config/parameter_reference.md#COM_DISARM_LAND))
## 예상 결과
- 추력을 올릴 때 서서히 이륙한다
- Drone should hold altitude in Altitude Flight mode without wandering
- Drone should hold position within 1 meter in Position Flight mode without pilot moving sticks
- 위에 언급한 어떤 비행 모드에서도 떨림이 나타나서는 안됨
- 지면에 착륙시, 콥터가 지면에서 튀면 안됨