Reliable eyes pose measurement for robotic bionic eyes with MEMS gyroscope and AKF filter

Abstract

In order to obtain the precise pose information of the robotic bionic eyes, the MEMS gyroscope is used for the measurement. Considering that the output of the MEMS gyroscope contains random noise, it is necessary to reduce the measurement error by noise reduction. According to the mathematical characteristics of the original output signal, the autoregressive (AR) model is established, and an adaptive Kalman filter(AKF) is designed to process the data. The results show that this method effectively reduces the random drift of the MEMS gyroscope.