Attitude Estimation Using Inertial and Magnetic Sensors Based on Hybrid Four-Parameter Complementary Filter

A complementary filter is presented for the attitude estimation using inertial and magnetic sensors. The proposed filter is based on a new attitude representation, where 4-D parameters (a 3-D unit vector and one angle quantity) are used. The proposed representation has an intuitive interpretation and the proposed filter has a separation property (magnetic disturbance does not affect pitch and roll angle estimation). Kinematics of the 4-D parameters and other properties are investigated. The proposed filter is verified through simulation and experiments.