Complementary Filter for Orientation Estimation: Adaptive gain based on dynamic acceleration and its change

An attitude and heading reference system that comprises a low-cost inertial measurement unit and a magnetometer is often inaccurate during highly dynamic motion. To mitigate sensor errors, we used a well-known complementary filter. The basic objective in using this filter was to compensate for the drift in the gyro by using the accelerometer and magnetometer as an assistance component. This paper presents the design of an adaptive mechanism to adjust filter gain using a fuzzy logic controller. We hypothesize that dynamic acceleration and change in dynamic acceleration can be used as inputs for the controller. The controller thus produces two adaptive gains for use as filter gains. The experimental result shows that the estimated angle has a good trend until a specific time, i.e., until the influence of the gyro-drift causes estimated angle divergence.