IMU array fusion algorithm based on improved Lie group extended Kalman filtering

• Published in: Youth Academic Annual Conference of Chinese Association of Automation (Online) pp. 404 - 408
• Main Authors: Chen, Yihao, Liu, Jieyu, Wang, Lixin, Xu, Jieyang, Shen, Qiang, Gao, Yongjia
• Format: Conference Proceeding
• Language: English
• Published: IEEE 17.05.2025
• Subjects: Accuracy; Approximation algorithms; Arrays; Data fusion; Data integration; Heuristic algorithms; IMU array; Kalman filters; Li Qun Extended Kalman Filter; Lie groups; real-time performance; Real-time systems; Simulation; Small angle approximation; Trajectory
• ISSN: 2837-8601
• DOI: 10.1109/YAC66630.2025.11150097

Inertial Measurement Unit (IMU), as an important sensor for acquisition of attitude and motion information, has a wide range of applications in the field of multi-sensor fusion. Since a single IMU is limited by sensor noise, drift and other problems, its measurement accuracy is often difficult to meet the demands of highly dynamic scenes. In order to improve the accuracy and real-time performance of IMU array data fusion, this paper proposes a fusion algorithm based on the improved Lie Group Extended Kalman Filter (LG-EKF). The proposed algorithm reduces the computational complexity by simplifying the Lie group mapping operation with small angle approximation, while retaining the advantages of LG-EKF in dealing with nonlinear systems. The superiority of the improved algorithm in terms of accuracy and real-time performance is verified by comparing the simulation results of 16 IMUs with real trajectories. The experimental results show that the algorithm significantly outperforms the EKF and is comparable to the accuracy of the LG-EKF in the estimation of the velocity, while significantly reducing the computational delay in the real-time computation, which provides an effective new scheme for the fusion of the multi-IMU data in the high-dynamic scenarios.