Hierarchical Fusion Algorithm for Multisensor Integrated Navigation System

衣晓,何友,关欣

DOI: https://doi.org/10.3321/j.issn:1000-1328.2003.02.008

2003-01-01

Abstract:Considerable research has been undertaken in the field of estimation theory in relation to the integrated navigation problem,which is of interest in both military and civilian applications.So far,the technique to do this has been using the federated filtering,a two level distributed fusion architecture first developed by Carlson.However,the multi level fusion architecture,another important multisensor information fusion mode,has not been applied to the integrated navigation system as far as know.The objective of this work is to introduce the hierarchical fusion algorithm to the integrated navigation system.Special attention is given to survey the difference between the hierarchical fusion algorithm and the federated filtering algorithm.The condition that the hierarchical fusion algorithm equals to the federated filtering algorithm is given in the form of the theorem.Then the part feedback mode of the hierarchical fusion algorithm is presented.The simulation results show that consecutive hierarchical fusion algorithm based on the part feedback mode is effective for integrated navigation system.It,with equal fault tolerant ability and computational complexity,can get better results than those of the federated filtering algorithm.