On FOG Strap-Down Attitude Heading Reference System Denoising and Design
|School||Harbin Institute of Technology|
|Course||Control Science and Engineering|
|Keywords||AHRS system FOG Random error Wavelet de-noising|
With the development of inertial devices and computer -based FOG AHRS system for its good stability, low cost , widely used in a variety of low accuracy of military and civilian fields . First , the issues the the AHRS system works and AHRS solver algorithm, and analyzes the system's main source of error . Systematic errors from the gyro noise , the FOG noise sources include quantization noise , angle random walk , bias instability , the speed of the random walk , speed slopes and Allan variance analysis of these noise sources . Wavelet analysis is a multi-resolution analysis ability to time-frequency analysis , suitable for non - stationary signal analysis . In this paper, wavelet theory and wavelet filtering method, by Allan variance analysis of the effectiveness of wavelet filter the FOG signal-to-noise removal . In the simulation analysis , by contrast the wavelet denoising effect of the gyro signal in the case of the different wavelets different vanishing moments , different decomposition level , different threshold processing methods , design appropriate filtering scheme , and in different circumstances wavelet conclusions of the filter is superior to the low-pass filtering . The subjects of single-axis FOG PC/104 embedded computer and three Russian production structures AHRS system , in which FOG measurement vector angular velocity , PC/104 for the completion of the AHRS calculated , and real-time display of LCD and CRT vector gesture corner . PC/104 has small size, low power consumption , high reliability , compatibility and so on . Software using the C language , the authoring environment is Turboc 2.0 , the main achievement of the the FOG signal acquisition , wavelet denoising, and fourth-order Runge-Kutta method is based on quaternion attitude solver .