Application of Comprehensive State Adaptive Vector Filter Algorithm in Vehicle Navigation

Currently, navigation technology is transitioning from SINS/GNSS integrated systems to multi-source information fusion. Faced with complex and ever-changing working conditions, some sensors may experience performance degradation, leading to system performance fluctuations and an inability to continuously output accurate navigation information. To address this issue, this paper proposes a combined navigation scheme consisting of SINS/GPS/odometer/ barometer /magnetometer. This information fusion algorithm is based on the federated Kalman filter (FKF) architecture and innovatively introduces the vector allocation factor algorithm, proposing a Comprehensive State Adaptive Vector Filter algorithm (CSAVF). At the subsystem level, calculate the F-norm of the covariance matrix, obtain the dynamic allocation factor, and then extract the singular values of the observable matrices of each system to obtain the allocation factors of each state variable. By incorporating the above factors into the fusion framework, the adaptive performance of the system can be improved. Finally, an unmanned ground vehicle navigation experimental platform was built, and semi-physical simulation experiments and real-time sports car tests were conducted. The experimental results show that the designed multi-source fusion navigation system has high adaptive performance, can effectively reduce the negative impact of partial sensor performance degradation on the integrated navigation system, and has high engineering application value.