Determinig of an object orientation in 3D space using direction cosine matrix and non-stationary Kalman filter

This paper describes a method which determines the parameters of an object orientation in 3D space. The rotation angles calculation bases on the signals fusion obtained from the inertial measurement unit (IMU). The IMU measuring system provides information from a linear acceleration sensors (accelerometers), the Earth's magnetic field sensors (magnetometers) and the angular velocity sensors (gyroscopes). Information about the object orientation is presented in the form of direction cosine matrix whose elements are observed in the state vector of the non-stationary Kalman filter. The vector components allow to determine the rotation angles (roll, pitch and yaw) associated with the object. The resulting waveforms, for different rotation angles, have no negative attributes associated with the construction and operation of the IMU measuring system. The described solution enables simple, fast and effective implementation of the proposed method in the IMU measuring systems.