Federal adaptive filtering of restraining consecutive outliers algorithm in multi-source information integrated navigation system
Bing Hua, Weidong Chen, Guohua Kang, Jing Yang, Fengxi Wu · 2013
In multi-source information Integrated Navigation System, the impact of outliers on the measurements information must be considered. First, the federated filter based outlier identification and elimination methods were analyzed for inertial, GNSS, astronomy and other types of multi-source information integrated navigation system. After outlier is removed, the measurement vector is reconstructed. Federal adaptive filtering algorithm with restraining continuous outliers algorithm is proposed. The algorithm is carried out through correcting the federal filtered local filter gain matrix and information distribution coefficient. Finally, simulation experiments are performed to verify the validity of the method. It can be observed that the impact of outliers on the precision of the filter is inhibited, thereby reducing the impact of consecutive outliers on the stability output of the system.