Integrated Attitude Estimate Algorithm for INS/Magnetometer System Based on Data Fusion of Dual MIMU Inertial Array

• Published in: IEEE sensors journal Vol. 25; no. 13; pp. 25410 - 25419
• Main Authors: Xue, Liang, Lu, Jixiang, Cai, Guangbin, Yang, Bo, Wang, Xinguo, Chang, Honglong
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.07.2025; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Arrays; Attitude estimation; Autonomous aerial vehicles; data fusion; Data integration; drift error; Geomagnetic field; Gyroscopes; inertial measurement array; Inertial platforms; Inertial sensing devices; Inertial sensors; Kalman filters; Magnetic field measurement; Magnetometers; Micromechanical devices; Noise measurement; quaternion; Quaternions; Sensor arrays; State vectors; Unmanned aerial vehicles; Vectors
• ISSN: 1530-437X; 1558-1748
• DOI: 10.1109/JSEN.2025.3566596

In this article, a micro inertial measurement unit (MIMU)/magnetometer attitude estimate algorithm based on the data fusion of an inertial measurement array is proposed for the micro unmanned aerial vehicle (UAV). A parallel and co-directional measurement array is constructed by using two MIMU units, and a Kalman filter (KF) is presented for fusing multiple gyro signals from MIMU array to achieve an optimal estimation of the tri-axial input rate, and then the estimated rate signals from fused MIMU are employed to resolve the aircraft's attitude quaternion. Furthermore, the quaternion error and drift noise of inertial sensors are chosen to construct a KF state vector, and an integrated KF is established to obtain the optimal attitude quaternion, specially a modified Gauss-Newton iterative approach is utilized to get the matched quaternion making use of gravitational and geomagnetic field information to construct the KF measurements. The simulations and experiments demonstrated that the estimated error of attitude quaternion using fusion of MIMU array is less than that by a single MIMU. The estimated errors for quaternion components of <inline-formula> <tex-math notation="LaTeX">{q}_{{0}} </tex-math></inline-formula> and <inline-formula> <tex-math notation="LaTeX">{q}_{{3}} </tex-math></inline-formula> are about 9.4% of the errors caused by the original signals of single MIMU, and the estimated errors for component <inline-formula> <tex-math notation="LaTeX">{q}_{{1}} </tex-math></inline-formula> and <inline-formula> <tex-math notation="LaTeX">{q}_{{2}} </tex-math></inline-formula> are about 15.3% and 17.3%, respectively. The results show that the proposed KF algorithm can effectively decrease the drift noise of MIMU and improve the attitude estimation accuracy.