Improved strong tracking Sage-Husa adaptive algorithm for multi-MEMS IMU data fusion

A circuit array of 16 micro-electro-mechanical system inertial measurement unit (IMUs) is developed, and an improved multi-IMU data fusion method based on the strong tracking Sage-Husa adaptive Kalman filter (ST-SHAKF) is proposed to achieve high-precision inertial measurement at low cost. The traditional Sage-Husa adaptive (SHAKF) algorithm is simplified for adaptive parameterization, with improved measurement noise variance estimation to ensure positive-definiteness. Filter divergence is addressed by supplementing the SHAKF with a strong tracking filter to maintain convergence. Dynamic weight allocation via minimum variance estimation enables effective multi-IMU data fusion. Experiments show that the proposed method significantly outperforms the traditional Sage-Husa adaptive Kalman filter in terms of Allan variance and standard deviation. Compared to traditional SHAKF, the proposed method achieves better noise suppression and improved fusion accuracy for both acceleration and angular velocity under both static and dynamic conditions.