An innovative transfer alignment method based on federated filter for airborne distributed POS

Abstract

The real-time and high-precision transfer alignment between the master and slave systems is the key issue for the airborne distributed Position and Orientation System (POS) to realize real-time and high-precision measurement of multipoint motion parameters. To solve the problem, this paper proposes a transfer alignment method based on federated filter, in which the high-dimensional state vector is divided into two parts and as the state vector of two local filters respectively instead of estimating in a centralized filter. This idea reduces the dimension of the system state vector; thus the computation speed is improved. The results of simulation and static ground experiment show that the proposed method runs faster than centralized filter by more than 10% and its alignment precision has the same level as centralized filter’s.