Design of speed calibration and estimation method for UAV MEMS gyroscope

The velocity of UAVs relative to the ground is usually measured using MEMS gyroscopes. In practice, the output speed of MEMS gyroscopes not only includes the true velocity of UAVs, but also includes random white noise and outlier points. Therefore, how to better process the velocity data measured by MEMS gyroscopes, remove white noise and outliers in the data, is of great significance for improving velocity accuracy and providing good data for velocity estimators[1]. Therefore, in order to obtain more accurate speed information, this chapter first performs denoising processing on sensor data based on wavelet transform, analyzes the speed measurement error of MEMS gyroscopes, and then uses adaptive identification methods to correct the installation deviation angle of MEMS gyroscopes[2]. In response to the random error of MEMS gyroscopes, an improved Sage Husa adaptive Kalman filtering method is used to estimate the true speed of UAVs, thereby obtaining accurate speed information. Simulation experiments have verified that this method has achieved good results.