The design of a kind of multi-wavelet signal filtering algorithm and its application to precision optical tracking servo system

Yang Pu, Ni Jiangfan and Li Xi

Abstract

This paper is aiming at non-stationary characteristics of gyroscope noise in precision optical tracking servo system, using the advantages of orthogonal wavelet transform, to present a multi-wavelets fast filtering method which is based on an energy weighted threshold, it can simplify multi-wavelets preprocessing method, and automatically change filtering threshold of different wavelet coefficients on the same scale through the design of a self adjusting energy weighted threshold, optimize multi-wavelets signal filtering algorithm. The practical application to optical servo system shows, this filtering method can take obvious inhibitory effect on gyroscope noise, the filtering effect is superior to Kalman filter which is based upon model and single wavelet filtering method with fixed threshold..