AeroBrainSTM

Autonomous STM32H743ZI Drone System

Hardware Requirements

• STM32H743ZI microcontroller board
• GPS module (e.g., NEO-6M)
• LIDAR sensor (e.g., TFMini Plus)
• OV7670 camera module
• Telemetry module (e.g., XBee)
• IMU (Inertial Measurement Unit)
• Drone frame and motors

Software Requirements

• STM32CubeIDE (version 1.8.0 or later)
• ARM GCC Toolchain
• OpenOCD for flashing

Key Features

1. Flight Control: Implements PID-based stabilization for roll, pitch, and yaw.
2. GPS Navigation: Utilizes GPS data for waypoint navigation.
3. Obstacle Avoidance: Uses LIDAR data to detect and avoid obstacles.
4. Path Planning: Generates and follows optimal paths between waypoints.
5. Telemetry: Enables real-time data transmission and remote command reception.
6. Computer Vision: Incorporates image processing for scene analysis and decision-making.

Build Instructions

1. Clone the repository:

git clone https://github.com/yezzfusl/AeroBrainSTM.git

cd AeroBrainSTM

2. Build the project:

make

3. Flash the binary to the STM32H743ZI:

openocd -f interface/stlink.cfg -f target/stm32h7x.cfg -c "program build/autonomous_drone.elf verify reset exit"

Usage

1. Power on the drone system.
2. Use the telemetry interface to:

• Add waypoints
• Start the mission
• Monitor drone status
• Abort the mission if necessary

Technical Details

• Microcontroller: STM32H743ZI (Arm Cortex-M7, 400 MHz, 2 MB Flash, 1 MB RAM)
• GPS: UART communication, NMEA sentence parsing
• LIDAR: I2C or UART communication, distance measurement up to 12m
• Camera: OV7670 module, DCMI interface, 640x480 resolution
• Telemetry: UART communication, custom protocol for data exchange
• Image Processing: Basic edge detection using Sobel operator
• Path Planning: A* algorithm for optimal path generation
• Obstacle Avoidance: Reactive avoidance based on LIDAR readings

Performance Considerations

• The main control loop runs at 100 Hz (10ms cycle time).
• Image processing is computationally intensive and may impact overall system performance.
• The system uses DMA for efficient data transfer from peripherals.

Safety Features

• Geofencing to restrict flight area
• Failsafe routines for loss of GPS or telemetry signal
• Low battery voltage detection and auto-return

Future Improvements

• Implement advanced computer vision algorithms for object detection and tracking
• Enhance path planning with dynamic obstacle avoidance
• Add support for multi-drone coordination

License

This project is licensed under the MIT License - see the LICENSE file for details.

Contributors

• [@YEZZFUSL]