Electrostatic Quantum Inspired Path Optimization (Autonomous Navigation) for UAVs. This project is still under further refinement and optimization.
This project implements an integrated UAV navigation system that combines:
- TENG sensor data processing for terrain mapping via triboelectric effects.
- A sliding electrostatic grid map construction.
- Quantum-inspired global path planning.
- Pigeon-inspired local path refinement.
- MAVLink communication with a Pixhawk flight controller.
The system is designed to run on a Raspberry Pi as a companion computer, interfacing with a Pixhawk flight controller (e.g., Pixhawk 2.4.8). It includes both simulation capabilities for testing and development, as well as support for real hardware deployment.
- Real-time TENG sensor data acquisition and calibration.
- Probabilistic grid mapping of terrain safety.
- Quantum-inspired candidate path generation and selection.
- Pigeon-inspired optimization for local path refinement.
- Simulation of UAV movement with battery and safety monitoring.
- MAVLink integration for flight control and telemetry.
- Extensive logging and diagnostics for development and debugging.
To run this project, you will need:
- A Raspberry Pi (tested with Raspberry Pi 5 8GB).
- A Pixhawk flight controller (e.g., Pixhawk 2.4.8) for real hardware deployment.
- TENG sensors connected via ADC channels for real hardware.
- Python 3.7 or higher.
- Required Python packages:
numpy,scipy,pymavlink(for real hardware). - For simulation: No additional hardware is required.
-
Clone the repository:
git clone https://github.com/elpantherd/EQIPO.git pip install -r requirements.txt
Note: For real hardware, ensure
pymavlinkis installed. For simulation, it is optional. -
Set up hardware (if using real UAV):
- Connect the Raspberry Pi to the Pixhawk via serial (e.g.,
/dev/ttyAMA0). - Connect TENG sensors to ADC channels on the Raspberry Pi.
- Connect the Raspberry Pi to the Pixhawk via serial (e.g.,
To run the EQIPO navigation system in simulation mode:
python eqipo.py --log_level INFOFor real hardware, ensure the correct serial port is set in the code and uncomment hardware-specific lines (e.g., GPIO, smbus, pymavlink imports).
--num_paths: Number of candidate paths to generate (default: 50).--grid_width: Grid map width in cells (default: 20).--grid_height: Grid map height in cells (default: 20).--log_level: Logging level (DEBUG, INFO, WARNING, ERROR).--log_file: Path to log file (optional).--sim_duration: Simulation duration in seconds (default: 60.0).
The system can be configured via global parameters in the code, including:
- ADC settings (e.g., number of channels, read interval).
- Calibration constants for TENG sensors.
- Grid dimensions and cell size.
- Optimization parameters for path planning and refinement.
Adjust these parameters in the code as needed for your specific setup.
-
Run with a custom grid size:
python eqipo.py --grid_width 30 --grid_height 30
-
Enable debug logging and save to a file:
python eqipo.py --log_level DEBUG --log_file debug.log
This project includes simulation capabilities for testing and development without hardware. In simulation mode:
- TENG sensor readings are simulated.
- MAVLink communication is mocked.
- UAV movement is simulated with battery drain and safety checks.
For real hardware:
- Uncomment hardware-specific lines in the code (e.g., GPIO, smbus, pymavlink imports).
- Ensure the correct serial port and baud rate are set for the MAVLink connection.
- Calibrate TENG sensors and adjust parameters accordingly.
- Sensor readings are noisy: Adjust calibration factors or increase grid smoothing sigma.
- Path planning fails: Ensure the grid map has sufficient safe areas or increase the number of candidate paths.
- MAVLink connection issues: Verify the serial port and baud rate in the connection string.
This software is proprietary and not open source. Please see the LICENSE file for terms of use.
For questions or support, contact TEAM QUASARS at [dthayalan760@gmail.com].