On SINS/GPS/GM Integrated Navigation Systems

As navigation techniques are developed, the performance of navigation systems is supposed to be better and better. Various integrated navigation systems have been important research directions, and that how to integrate various navigation equipments and take full advantage of redundancy to get a integrated navigation system with high accuracy and high fault tolerance is full of practical significance. With a navigation system which is applicable to high-altitude unmanned reconnaissance aircrafts as background, a SINS/GPS/GM integrated navigation system is studied. Based on the research of each local navigation system and information fusion theory, a SINS/GPS/GM integrated navigation algorithm is designed.Inertial navigation systems and global positioning systems have different traits. Integrating SINS/GPS and complementing their performance is an important scheme for highly autonomous and accurate navigation. SINS/GPS integrated navigation systems can give out accurate position and velocity, but it can't give out accurate attitude. So GM which helps to measure attitude is introduced into the integrated navigation system. Attitude error measurement equation is established through the coordinate transformation between geomagnetic vector that measured by magnetometers and that computed in IGRF. With the small quaternion constituted by attitude errors, attitude quaternion update equation is presented, which is a more concise method for the update of attitude matrixes.As a special case of distributed filtering, the federated Kalman filtering algorithm is the core technique of navigation information fusion. So based on the discussion about structure traits, a non-feedback federated Kalman filter which consists of position sub-system, velocity sub-system and attitude sub-system is designed and simulated. In order to improve the adaptivity of main filter when the performance of local filters decline, an adaptive-information-fusion federated Kalman filtering algorithm based on eigenvalue decomposition of the estimated state covariance matrix, is studied. Further more, an improved adaptive-information-fusion method which is based on breakdown state of each local filter is presented. The simulation results show that the improved adaptive-information-fusion method makes the integrated system have higher accuracy.In order to improve the fault-tolerance of the SINS/GPS/GM integrated navigation system, the theory of fault-tolerant design for integrated navigation systems is studied.χ2-detection method for validity of local filters measurement data is mainly studied. A fault-detection-isolation-reconstruction algorithm for local filters is designed based on Residual-χ2-detection method in this SINS/GPS/GM integrated navigation system. The simulation results show that the fault-tolerance of the integrated navigation system is improved.