Real-Time UAV Attitude Heading Reference System Using Extended Kalman Filter for Programmable SoC

摘要

We implemented an extended Kalman filter (EKF) hardware for a real-time attitude heading reference system (AHRS). Attitude estimation is an important process for small unmanned aerial vehicle (UAV) control. Small UAVs mounts smaller sensors due to the limitation of space and power. However, effects of noise components in sensor output become large if the size of sensors become smaller. We considered using EKF to reduce effects of noises by sensor fusion. Moreover, real-time and highly accurate attitude estimation is required for an autonomous flight of UAVs. To reduce loads of processors in a real-time calculation of EKF, we implemented a circuit of EKF by C-based hardware design in Programmable SoC. We achieved five times faster EKF processing on FPGA than the processing on ARM Cortex-A9 @677Hz with single core processing.