A Kalman Filter Algorithm for Attitude Estimation in Mems Imu with Presence of Disturbance Caused by External Acceleration

Abstract

This paper presents a Kalman filter algorithm for attitude estimation in MEMS IMU, which is robust in the presence of unexpected or external acceleration. The algorithm decomposes the true attitude direction cosine matrix (DCM) into estimated and error components.  to improving attitude estimation accuracy. Additionally, within the error state Kalman filtering model, a strategy based on model compensation and measurement data monitoring is designed to adaptively adjust the measurement noise matrix online. This strategy aims to mitigate the impact of disturbance caused by external acceleration on attitude estimation, enhancing the algorithm's measurement accuracy and robustness under dynamic conditions. Finally, we validate the effectiveness and measurement accuracy of the proposed algorithm through laboratory rotation experiments and land vehicle tests.

Keywords: Attitude estimation, low-cost MEMS IMU, robust adaptive, error-state Kalman filter (ESKF), disturbance acceleration compensation.