Improvement of Federal Filtering Algorithm for Vehicle Integrated Navigation System
Shu Dong-liang · Measurement & Control Technology · 2008
Precision and reliability are important performance indexes for vehicle integrated navigation system,conventional no-reset federal algorithm based on output correction has the best fault-tolerance performance,but error divergence as long time navigation.Improving no-reset federal filtering algorithm based on local feedback,studying the problem of weight matrix construction for local filtering and assuring,the isolation of two local filteri's,each other,optimal reliability can be achieved,and local feedback can advance the precision of filtering.Three vehicle navigation experiment results prove that,the higher position precision of improvement algorithm is better than that of no-reset federal algorithm based on output correction.