Project brief
A strapdown Inertial Navigation System using tactical-grade MEMS accelerometers and gyroscopes, fused with dual-constellation GNSS via an Extended Kalman Filter. The system achieves sub-meter accuracy in GPS-aided mode and maintains reliable dead-reckoning for 90 seconds during GNSS outages. The EKF runs at 200Hz on a dedicated STM32H7 processor, with ARINC 429 output to cockpit displays.
This project brings sensor calibration, embedded estimation, and avionics output into one navigation pipeline. Any accuracy figures remain claims in the supplied record rather than newly verified results.
The engineering challenge
Maintain a coherent navigation estimate while inertial errors accumulate and satellite aiding becomes intermittent.
Engineering approach
- Represent inertial bias and navigation states in an Extended Kalman Filter.
- Apply the stated temperature compensation and alignment modes.
- Publish fused attitude, position, and velocity over the avionics output interface.
Features & capabilities
- 9-DOF MEMS sensor fusion — accelerometer, gyroscope, magnetometer
- Extended Kalman Filter at 200Hz update rate
- <0.1° roll/pitch/heading accuracy in GPS-aided mode
- 90-second GPS outage dead-reckoning with <50m CEP drift
- Dual-constellation GNSS (GPS + GLONASS) for improved availability
- ARINC 429 output: attitude, heading, position, velocity, acceleration
- Built-in alignment modes: gyrocompass, stored heading, GPS heading
- Temperature compensation across -40°C to +85°C operational range
Software & engineering tools
STM32H743 (Cortex-M7, 480MHz), ADIS16497 tactical IMU, u-blox M9N GNSS, FreeRTOS, C++17, Altium Designer