Inertial Sensor Error Model Selection Based on Dominant Energy Features

Li Fu, Kanli Tian, Lingling Wang

Abstract:	In practical application, the accuracy of SINS is limited as a result of errors of inertial sensors. In order to improve the performance of SINS, the error models have been built to predict the performance of inertial sensors for particular circumstances. The ideal error models of inertial sensors should contain all effects which give rise to errors in measurements. However, the ideal models with multi-effects are usually sophisticated and high dimension. Furthermore, for multi-effects analysis and compensation, it is more complex and difficult to implement when the high order error model is considered. In this paper, error models of inertial sensor of SINS for monitoring and predicting, which take care of temperature changes, are proposed. The real time data of various variables (namely, temperature, the rate of temperature change, temperature gradient, the input angular rate and gyro output) are collected in laboratory environment. MANOVA tests has been carried out to know the quantitative measure of the different temperature. PCA and PLS regression methods are applied on the obtained data sets to monitor and predict the output of gyro. Finally, the results of this study show that the error models can predict the future data of gyro to reach the aim of accurate compensation.
Published in:	Proceedings of the 27th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2014) September 8 - 12, 2014 Tampa Convention Center Tampa, Florida
Pages:	1807 - 1811
Cite this article:	Fu, Li, Tian, Kanli, Wang, Lingling, "Inertial Sensor Error Model Selection Based on Dominant Energy Features," Proceedings of the 27th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2014), Tampa, Florida, September 2014, pp. 1807-1811.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In