Tune Position and Yaw Setpoints in External Mode and Log Flight Data Using PX4 SIH

Prerequisites

1. If you are new to Simulink, watch the Simulink Quick Start video.

2. If you have not done so already, install the UAV Toolbox Support Package for PX4 Autopilots. Select the installation topic appropriate for your operating system.

At the Select a PX4 Autopilot and Build Target step of installation, specify these options::

For more information on board and the build targets, see SIH on Flight Controller Hardware.

3. Before running Software-In-The-Hardware (SIH), add the following line to the default PX4 board configuration file for the target hardware:

CONFIG_MODULES_SIMULATION_SIMULATOR_SIH=y

Save the updated .px4board file. Then in the Hardware Setup screen select the file using the Custom PX4Board File option, and use the multicopter CMake variant. For more information about modifying PX4 build target files and enabling or disabling PX4 modules see Disabling Modules in PX4board Build Target File. Then run the hardware setup screen with the multicopter variant of cmake.

4. In QGroundControl (QGC), configure the following for SIH operation:

5. Install Mission Planner. This example uses Mission Planner to configure the telemetry radio modules required for wireless External Mode communication.

Required Hardware

PX4 Simulation-In-Hardware (SIH)

PX4 Simulation-In-Hardware (SIH) is a simulation mode in which the PX4 firmware runs a built-in vehicle dynamics model directly on the autopilot hardware. Unlike Software-in-the-Loop (SITL), which runs the PX4 firmware and a dynamics model on a development computer, SIH executes entirely on real autopilot hardware and uses the internally simulated vehicle state instead of physical sensor data.

SIH enables you to verify flight controller algorithms on the actual target hardware without requiring:

SIH is useful for iterating on controller designs because you work directly with the hardware that runs your deployed code, while the firmware's internal dynamics model provides realistic vehicle dynamics feedback.

Note: The built-in SIH dynamics model is not available on all PX4 hardware targets due to memory constraints on some boards. Verify that your PX4 hardware supports SIH before using this example. PX4 Pixhawk 6c and similar boards with sufficient flash memory support SIH by default.

Model Overview

Open the WirelessExtModeSIH.slx model.

modelName = "WirelessExtModeSIH";
open_system(modelName)

The model consists of these components:

Model Initialization Script

The model uses an initialization script, initializeModel.m, to define the sample time and PID controller gains for the position, velocity, altitude, and yaw controllers. Run this script before deploying the model to load the required parameters into the MATLAB workspace.

initializeModel

The script sets platform-specific PID gains to account for timing differences between Windows and Linux development environments.

Position and Altitude Controller

The Position & Altitude Controller subsystem implements a cascaded control structure with an outer position loop and an inner velocity loop.

Outer position loop:

Inner velocity loop:

Attitude Controller

The Attitude Controller subsystem computes torque commands to track the desired roll, pitch, and yaw angles.

The desired roll and pitch angles come from the position controller output. The desired yaw angle comes from the input slider.

Command to Actuators

The To Actuator subsystem maps the thrust and torque commands to individual motor speeds using a mixer matrix for an X-configuration quadcopter.

The mixer matrix distributes the collective thrust and the roll, pitch, and yaw torques to the four motors. The motor commands are converted to single precision before being sent to the PX4 actuator output block. The Arm signal controls whether the motors are active.

Configure Telemetry Radio Modules

Wireless External Mode requires a pair of telemetry radio modules (air and ground). Configure both modules to use the same baud rate as the External Mode communication channel.

To configure the telemetry modules:

Note: Both telemetry modules must use the same baud rate. The baud rate must also match the value configured for External Mode communication in the Simulink model. Common baud rate values are 57600 and 115200.

Configure External Mode Communication in Simulink

After the telemetry modules are configured, set up the Simulink model to communicate over the serial link. These settings are pre-configured in the model. Update the serial port and baud rate to match your telemetry hardware.

3.Select the Hardware board serial port to the serial port where the air module is connected (for example, /dev/ttyS4). For the correct serial port name, see Serial Port Names and Corresponding Labels on PX4 Flight Controller Boards.

4. Set Baud rate (in bits/s) to the same value configured on the telemetry modules (for example, 57600).

5. If listed, clear the Use the same host serial port for External mode as used for firmware upload option.

6. Set Host Serial port to the COM port on the development computer where the ground telemetry module is connected (for example, COM9).

7. Click Apply and Ok.

Deploy and Run the Controller in External Mode

To deploy the model and run it in external mode with SIH:

To run the quadcopterPositionController.slx model in Monitor & Tune mode:

Simulink builds the model, deploys the generated code to the PX4 hardware, and establishes the wireless external mode connection.

After the connection is established, the PX4 firmware uses its built-in SIH dynamics model to simulate vehicle behavior. The simulated vehicle responds to your controller commands.

Tune Position and Yaw Setpoints

During external mode operation, use the dashboard slider blocks to command the vehicle to different positions and yaw angles:

Adjust the sliders and observe the real-time vehicle response in the Display and Scope blocks within the model. The controller tracks the commanded setpoints using the cascaded PID architecture.

Configure uLOG Logger for Simulink Custom Message

Before you can log controller signals to the PX4 SD card, register the custom uLOG message topic in the PX4 logger configuration. The PX4 logger reads a file named logger.txt on the SD card to determine which additional topics to record.

To configure the logger:

simulink_custom_message

After this configuration, the PX4 logger records the simulink_custom_message topic alongside standard flight data whenever logging is active. You only need to perform this step once; the configuration persists on the SD card across power cycles.

For more information on uLOG logging, see Log Simulink Signals Using PX4 ULog.

Log Flight Data Using uLOG

Logging enables you to evaluate how well the multicopter follows commanded setpoints and identify areas for controller tuning. The model logs flight data to the PX4 SD card using the uLOG logging mechanism with the simulink_custom_message topic.

The logged fields in the simulink_custom_message topic include:

Note: Ensure that you have configured the simulink_custom_message topic in the logger.txt file on the SD card before the flight session, as described in the previous section. \

To retrieve and analyze the flight log:

Analyze Tracking Performance and Tune PID Gains

After you retrieve the logged data, read the simulink_custom_message topic from the uLOG file and compare the desired and actual position values to evaluate controller performance.

Plot the desired and actual position for each axis:

% Read the uLOG file
logData = ulogreader("log_file.ulg");

% Extract Simulink custom message data
sihData = readTopicMsgs(logData,"simulink_custom_message");

% Convert timestamps to seconds (start from t = 0)
t = (sihData.timestamp - sihData.timestamp(1)) * 1e-6;

% Plot X position: desired vs actual
figure
subplot(2,1,1)
plot(t, sihData.x_actual, 'b', 'DisplayName', 'Actual X')
hold on
plot(t, sihData.x_desired, 'r--', 'DisplayName', 'Desired X')
xlabel('Time (s)')
ylabel('X Position (m)')
legend
title('X Position Tracking')
grid on

% Plot Y position: desired vs actual
subplot(2,1,2)
plot(t, sihData.y_actual, 'b', 'DisplayName', 'Actual Y')
hold on
plot(t, sihData.y_desired, 'r--', 'DisplayName', 'Desired Y')
xlabel('Time (s)')
ylabel('Y Position (m)')
legend
title('Y Position Tracking')
grid on

From the plots, you can analyze

Other Things to Try