An EMD-MRLS de-noising method for fiber optic gyro Signal

Abstract

In order to reduce the influence of fiber optic gyroscope (FOG) random drift error on inertial navigation systems (INS), a novel de-noising method is proposed in this paper, which based on an improved empirical mode decomposition (EMD) and modified recursive least square (RLS). Referred to as the EMD-MRLS method, it is developed to decompose the FOG outputs into a number of intrinsic mode functions (IMFs) after which mode manipulations are divided into three parts, noise IMFs, mixed IMFs, and information IMFs by two index parameters based on mahalanobis distance (MD). A modified RLS algorithm is then employed to process the mixed IMFs, from which the refined IMFs components are reconstructed to produce the final de-noising results. Other traditional methods, such as RLS, and EMD are investigated to provide a comparison with the proposed one through both simulated signals and experimental FOG outputs. Compared with the EMD method, the results show that the error mean is reduced by 27.01%, and the horizontal position error is reduced by 106.75m, when the INS lasts for 1000s.