Comparing two methods of coarse self alignment using high accuracy SINS

Leonardo Kameyama de Castro Leite¹; Helio Koiti Kuga²; Waldemar de Castro Leite Filho¹

¹ INPE - Brazilian Institute for Space Research; ² ITA - Technological Institute of Aeronautics

doi:10.20906/CPS/CILAMCE2017-0553

Resumo

Initial alignment is a classical problem existing when one needs to bootstrap an inertial navigation system (INS) based on inertial sensors such as gyros and accelerometers. When no external aid is present one needs to provide the INS with some self-starting procedure. This study performs an investigation on the performance of two methods of coarse self-alignment for Strap-down Inertial Navigarion Systems (SINS). The probably best-known standard method TRIAD is compared to a new variant method of coarse self-aligment known as ON-TRIAD1. The performance assessment between the two methods is given by the comparison between the Euler angles yielded by each method, using actual data from relatively high accuracy SINS (aerospace grade). Using the initial static part of high sampling data, both methods are exercised and the Euler angles are computed several times. Is was also analyzed the influence on the accuracy of a simple pre-processing scheme to identify gross outliers, with little additional computer burden. In short, this work describes the results of both procedures depicting the influence of different mathematical tools in the data processing, accounting for the measurement quality as well.

Palavras-chave: Aided navigation; Inertial navigation; Coarse alignment