On FOG Strap-Down Attitude Heading Reference System Denoising And Design

With the development of computers and inertial sensors, the strap-down attitude heading reference system (SAHRS) based on fiber optic gyroscope (FOG) has been widely applied in military and civil fields with medium or low accuracy because of its characteristics of high reliability and low cost.Firstly, the basic concepts and principle of the strap-down attitude heading reference system and the algorithms for computing attitude angles are introduced, and then the main system errors, which mostly come from FOG noises are analyzed. The main FOG noises are composed of quantization noise, angle random walk, bias instability, rate random walk and rate ramp. And their coefficients were estimated through Allan variance method.Wavelet analysis is new tool for time-frequency analysis with multi-resolution, that is suitable for analyzing non-stationary signal. The basic concepts and theories of wavelet analysis and wavelet de-noising are introduced, and the de-noising effectiveness is certified through Allan variance method. In the process of wavelet threshold de-noising method simulation, different de-noising effectiveness, corresponding to different wavelet basis, different vanishing moments, different decomposition levels, different threshold rescaling methods, is analyzed. And so the author designs a proper wavelet de-noising program. Then the superiority of wavelet de-noising over low-pass filtering is certified.A strap-down attitude heading reference system based on three fiber optic gyroscopes made in Russia is built, and PC/104, which is an embedded computer with characteristics of small size, low power consumption, high reliability and good compatibility, is used to compute the three attitude angles as are showed by a set of LCD and a set of CRT in real time. The software, including gyro data-sampling term, wavelet de-noising term, term for computing three attitude angles with 4-Runge-Kutta method based on quaternion method, is implemented with C language in the environment of Turboc 2.0.