A feasibility study on attitude filtering without rate gyro

In this paper, an improved attitude determination system without rate gyro based on attitude angle differential of TRAID algorithm and extended Kalman filter (EKF) is presented. An improved observation structure with the attitude angle from TRAID algorithm and its differential both treated as observations is established. The observation error covariance matrix is derived based on two three-dimensional coordinates related to rotation angle errors of multi-vector measurement system. EKF is used to estimate the attitude angle and angular velocity. The observability and estimability of the nonlinear attitude estimation system are analyzed via the eigenvalue decomposition method and the singular value decomposition method. Numerical simulations are undertaken to verify the improved attitude determination method without rate gyro. The results illuminate that the improved attitude determination system is better observability and estimability than the traditional method and is robust to the initial attitude estimation error.