Research on MEMS Gyro Random Drift Restraining based on Simplified Sage-Husa Adaptive Filter Algorithm

摘要

The various random noises in MEMS gyro are the main factor affecting the accuracy. In accordance with the property of its noise, traditional filtering has many shortages. If so, a simplified Sage-Husa adaptive filter algorithm is discussed. The algorithm can estimate the system process noise accurately when observation noise is known. The paper sets up ARMA model of MEMS gyro random error and processes the drift data by the filtering. The result shows that the random noise is reduce by 70% .