Reseach On Signal Processing And Denoising Technique Of Fiber Optic Gyroscope

Inertial navigation system(INS) has been widely applied in national defense, aviation, aerospace and other domain. With the development of inertial technology, there is an increasing trend of INS that traditional mechanical gyros are replaced by new types of solid-state gyros and platform inertial navigation system is replaced by strap-down inertial navigation system. Along with the development of photoelectricity technology, optics sensors have become a new generation of ineria measurement unit (IMU). Depending on the advantages such as small in size, light in weight, easy to be vindicated and better cost performance, fiber optic gyroscope (FOG) becomes the preferred IMU in low or medium accuracy INS.Since the precision of navigation system is the key of its application, it is urgent and important to improve the precision of navigation systems based on inherent hardware. Filtering method based on software technology can save the cost and improve system accuracy. The usual solution is to use filtering method, namely signal filtering of measurement sensor output and state filtering of navigation system. Stochastic noise in FOG consists of white noise and fractal noise. Because fractal noise is non-stationary random process, it is difficult to eliminate fractal noise by traditional methods. Thus, one of the tasks in this study is to propose a new nonlinear filtering method for noise elimination of FOGs to improve the accuracy of FOGs, and to improve the accuracy of navigation system.On the existing filtering method of FOGs'random noise, with regard to noise source analysis and noise filtering method of different precision-level FOGs, have not yet reported deeply and widly at home and abroad. Therefore, lifting wavelet and fractal theory is introduced in this study to FOGs'signal processing and stochastic noise filtering, and this research is based on space technology research areas related to the status quo, so it is of great practical significance and theoretical value.In this research, the variation of stochastic noise's parameters and the influence of noise parameters to interferometric fiber optic gyroscope (IFOG) is studied. Combined with fractal theory, wavelet method, lifting wavelet approach and empirical mode decomposition method are reseached to eliminating the stochastic noise in FOG. This study combined with the fund project of Aerospace Science and Technology Innovation Foundation of China"Application of fiber optic gyro denoising based on wavelet and fractal theory"mainly has the following several aspects of work:1. Research on characteristic analysis, simulation and model of FOGBased on introducing the principle and structure of open / closed-loop FOG, the composition and characteristics of stochastic noise in FOGs are analysed, in addition, the dynamic and stochastic model is simulated and modeled. Comparative analysis of the noise in different precision-levels of FOG is implemented to provide theoretical guidance for the follow-up study of the noise analysis and filtering.2. Application of parameters estimation based on fractal theoryMaximum likelihood estimation method is introduced to estimate the parameters of stochastic noise in FOGs, and the derivation, inspection processes, and algorithms of MLE in wavelet transform domain is given. This noise parameter estimation approach has contributed to the performance evaluation of FOG, and to be a theoretical guidance to improve the accuracy of FOG. The comparison of estimation results with the actual performance of FOG verifies this method. Meanwhile, according to the structure of the FOG, the estimation of stochastic noise parameters is to guide to improve FOGs'accuracy from the composition of the device.3. Denoising method in FOG based on WTBasic theory of wavelet transform is introduced and the denoising method based on wavelet transform is discussed. Simulated experiments are researched, including decomposition and reconstruction of Mallat arithmetic, soft threshold wavelet denoising method, different threshold-selecting wavelet denoising method and the traditional method to filter the noise of FOG.4. Denoising method in FOG based on LWTAfter a brief introduction of the mathematical foundations of lifting wavelet analysis, an adaptive lifting wavelet denoising method based on several wavelet bases is proposed to eliminate the fractal noise and Guassian white noise, and experimental results prove the effectiveness of the method. The traditional wavelet denoising method is also studied as a comparison with new method, and the analysis of actual FOG data shows the validity of LWT denoising method.5. Denoising method in FOG based on EMDTo overcome the shortcomings of empirical mode decomposition (EMD) analysis——low resolution in high-frequency signal, a type of mixed EMD-LWT method is proposed and applied in elimination of stochastic noises for a FOG. Experimental results on real FOG data show that the proposed mixed method is feasible and effective in denoising. This research work can provide reference for signal processing and noise elimination of FOGs in inertial technology domain.