Error Estimation of Inertial Sensors using Allan Variance

Abstract

The demand for the high navigation accuracy and low power consumption for Unmanned Underwater Vehicle (UUV) has been increasing as per mission requirement. The Inertial Navigation System (INS) provides better solution viewing these necessities. Accelerometers and gyros are main components of an INS. Gyro plays key role in achieving navigation accuracy. Therefore it is very important to model the gyro's intrinsic errors. Angle Random Walk (ARW) is the noise component perturbing the output of Fiber Optic Gyro (FOG). Allan Variance method is adopted to identify the ARW of particular FOG before and after thermal-inertial calibration. The reduction in ARW from 28% to 60% is observed experimentally.