Fixed-Wing Flight Controller

System Specifications

• Microcontroller: STM32L4A6ZG Nucleo (bare-metal C)

• Sensors: 6-DOF IMU (accelerometer, gyroscope)

• Filtering: Custom Kalman filter for pitch/roll/yaw estimation

• Control: PID loops for roll, pitch, yaw stabilization

• Actuators: PWM output to control ailerons, rudder, and elevator

• Telemetry: UART serial link, real-time PID tuning and flight data logging via MATLAB

• Power: Li-Po battery, voltage regulators for 3.3V and 5V rails

• Size: 324 mm × 96 mm × 50.3 mm

• Weight: 1,360.7 g

• Battery life: >18 hours

• Servos: SG90, 5 V

• IMU: MPU6050, 3.3 V, 1 kHz sample rate

• Interfaces: UART (115200 baud), I²C (100 kHz), TIM5 (1 MHz)

Design & Architecture

Results & Lessons Learned

Results

• SteadiFly successfully demonstrated real-time stabilization of pitch, roll, and yaw in a fixed-wing aircraft test rig.
  
  

• The Kalman filter smoothed noisy accelerometer/gyroscope data, producing stable orientation estimates.
  
  

• Independent PID controllers for each axis provided consistent corrective actuation through servos.
  
  

• The UART interface allowed in-flight PID tuning and MATLAB logging, which accelerated testing and parameter optimization.
  
  

Lessons Learned

• Sensor tuning is iterative: Calibrating gyroscope bias and tuning the Kalman filter required repeated test flights and log analysis.
  
  

• Low-level I²C configuration (without STM32 HAL) was challenging, but ultimately gave tighter control over timing and reliability.
  
  

• UART-based PID tuning avoided the need for constant code reflashing, saving hours of debug time.
  
  

• Servo mapping and scaling were critical — improper ranges caused instability until carefully constrained.
  
  

• Future improvements would include GPS integration for waypoint navigation, additional redundancy, and expansion toward a full autonomous flight controller.

Documentation