Study On The Signal Processing Of Real Time System Based On Wavelet Method

With the development of inertia technology, the required precision of the gyroscope has been much higher. Gyro drift is an important target that can affect the precision of gyroscope, so the research on gyro drift filtering is much more important. In this dissertation, the wavelet de-noising theory is applied in gyro drift signal processing and gets an inspiring result. In this dissertation, it makes ulterior study on wavelet de-noising methods.Firstly, the paper introduces fundamental of wavelet transform. Mallat algorithm of discrete wavelet is put forward based on introducing of sequential wavelet transform and discrete wavelet transform. It is discussed that Multi-Resolution Analysis on signal with wavelet transform. The paper introduces relation between wavelet transform and Lipschitz exponent and analyzes characteristic of signal and noise under wavelet analysis. Threshold de-noising method is introduced mostly at last.Secondly, based on the selection of the wavelet bases of gyro drift signal, bi-orthogonal wavelet is used in gyro signal processing systematically and gets an inspiring result. It provides experimental basis of wavelet de-noising applications on inertial navigation system. Wavelet de-noising method programmed with C language is used in inertial navigation system systematically and gets an inspiring result.In the end, in the ulterior study on wavelet de-noising methods, two kinds of new threshold de-noising methods, which are based on fuzzy wavelet de-noising methods, are firstly presented and have inspiring de-noising result. Ascertain method of decompose layer is put forward on the characteristic of wavelet decompose coefficient. Processing signal with gauss whit noise and impulsive noise utilizing of method put forward by the paper can get rid of gauss whit noise and impulsive noise and surge phenomena brought by traditional threshold de-noising.