Application for GPS/SINS loosely-coupled integrated system by a new method based on WMRA and RBFNN

Authors:
Xiyuan Chen;Xuefen Zhu;Zigang Li
Affiliations:
School of Instrument Science and Engineering, Southeast University, Nanjing City, P.R. China;School of Instrument Science and Engineering, Southeast University, Nanjing City, P.R. China;School of Instrument Science and Engineering, Southeast University, Nanjing City, P.R. China
Venue:
ICIC'07 Proceedings of the intelligent computing 3rd international conference on Advanced intelligent computing theories and applications
Year:
2007

Citing 4
Cited 1

The fast wavelet transform

Dr. Dobb's Journal
Neural networks for control systems: a survey

Automatica (Journal of IFAC)
Neural Networks: A Comprehensive Foundation

Neural Networks: A Comprehensive Foundation
Fast learning in networks of locally-tuned processing units

Neural Computation

Wavelet multi-resolution analysis aided adaptive Kalman filter for SINS/GPS integrated navigation in guided munitions

AICI'11 Proceedings of the Third international conference on Artificial intelligence and computational intelligence - Volume Part II

Quantified Score

Hi-index	0.00

Visualization

Abstract

A new non model-related algorithm that can perform the autopiloting of the aircraft under all conditions is presented. For improving the precision of the loosely coupled GPS/SINS integrated navigation system, fusing data from a SINS and GPS hardware utilizes wavelet multiresolution analysis (WMRA) and Radial Basis Function Neural Networks (RBFNN). The WMRA is used to compare the SINS and GPS position outputs at different resolution levels. These differences represent, in general, the SINS errors, which are used to correct for the SINS outputs during GPS outages. The RBFNN model is then trained to predict the SINS position errors in real time and provide accurate positioning of the moving aircraft. The simulations show that good results in SINS/GPS positioning accuracy can be obtained by applying the new method based on WMRA and RBFNN.