Optimization of gyroscope and accelerometer/magnetometer portion of basic attitude and heading reference system

摘要

An Attitude and Heading Reference System (AHRS) provides orientation information by fusing sensor data from a magnetic and inertial measurement unit (MIMU). A MIMU consists of three components: gyroscope, accelerometer and magnetometer. A basic AHRS approach fuses two complementary components together using a weighted approach. The first component estimates the relative orientation from the gyroscope, and the second component provides the absolute orientation from the accelerometer and the magnetometer. However, the weighting of these two components is critical to obtain the best orientation information. In order to find the optimal weight a genetic algorithm is used for the optimization task. Results obtained via ground-truth simulated MIMU signals with real noise content reveal that the RMSE orientation error (root means squared error) of the Euler angles in degrees can be reduced by a factor of 3.6 (worst RMSE = 10.054 and best RMSE = 2.764 obtained during a GA run).