Adaptive Complementary Filtering Algorithm for IMU Based on MEMS

摘要

Aiming at the inaccuracy of the attitude angle estimation caused by low measurement accuracy of MEMS sensors, this paper proposed an adaptive sparse interpolation lossless complementary filter (ASICF) based on quaternion and complementary filtering. Data fusion of the MEMS accelerometer and the MEMS gyroscope is performed in ASICF to calculate roll and pitch. By down sampling frequency down-conversion operation, the sampled data is adaptively preprocessed. ASICF is verified by the MAV dataset, the self-driving car dataset and the human pose dataset, and it is proven that the algorithm has high accuracy under the condition of large disturbance of acceleration or dramatic attitude change.