Study on Iterated Divided Difference Filtering and Its Application on TA Process with Large Initial Misalignment Angle

Abstract

With provision for minishing the influence of higher-order truncation error in DDFII algorithm on the state estimation problem of transfer alignment (TA) with large misalignment angles, it developed the Iterated DDFII (IDDFII) algorithm. This algorithm applied the estimation value of state vector to linearize the nonlinear system function more times based upon the latest observation vectors, and updated the last time estimated value of state vector to make it more accurately approaching to the true value of state vector better. With using the observation information more times, it decreased the influence of the higher-order truncation error of nonlinear system equations on state vectors estimation precision, and furthermore it analysized the limitation and application condition of IDDFII algorithm. At the same time based on the real-time and higher-precision filtering requirements of the TA system with large initial misalignment angles, it developed its' nonlinear system model, and studied this error model with the IDDFII and DDFII algorithms. The simulation results indicate that the IDDFII algorithm can primly meet the performance requirements of TA system, comparison to DDFII algorithm, IDDFII algorithm can improve the precision index of TA process on condition that it doesn't effect the real-time requirements.