Angular Motion Estimation using Distributed Accelerometers
Final Report Abstract
The main objective of this project was to develop different novel approaches for improving the performance of inertial sensors through deriving and developing novel sensor data fusion approaches for the angular motion estimation using distributed accelerometers’ configurations. From experimental and simulation results obtained throughout the project, the main objective of the research has been achieved. We focused on the use of distributed accelerometers for inferring the angular motion from the angular information contained in their measurements. We developed different fusion schemes for benefiting from the AIV to form a GF-IMU or to aid the GF-IMU by conventional gyros. Two novel solutions for estimating the angular motion from a multiple distributed tri-axial accelerometers measurements in a GF-IMU were presented and verified experimentally. The integration scheme was realized using an extended Kalman filter (EKF). One approach is by using direct Euler integration model and the other scheme utilizing dynamic models in 3D which is capable of estimating bias parameters in a GF-IMU. Furthermore, we derived nonlinear equality constraints and presented a filter scheme to benefit from the constraints in improving estimation performance. The benefits of using the hybrid IMU over the GF-IMU were illustrated. Furthermore, a novel configuration of rotating and fixed accelerometers to find the angular motion was developed and its governing equations were derived. This configuration was used to find the north for the static scenario. The main contributions of this work, with emphasis on their novelty, are listed as follows: • Development of a GF-IMU using a configuration of twelve fixed accelerometers and adopting a calibration procedure for the GF-IMU. This configuration benefits from the angular information vector (AIV) with a simple calibration procedure. • Deriving two Kalman models for angular motion estimation (AME) using the AIV. The first model is a direct model based on Euler integration. The second model is based on dynamic models and it is capable of estimating the bias vector in the AIV due to accelerometers’ bias vector. This is a novel solution to obtain a biasfree IMU. • Using distributed accelerometers to aid conventional gyros to avoid the drawbacks of both types of sensors and to get the best of both. • The development of a GF-IMU using fixed and rotating accelerometers configurations and applying this principle to construct a north finding system. • The development of a GF-IMU using fixed and rotating accelerometers configurations and applying this principle to construct a north finding system. • Developing a novel DCM based GPS/INS integration algorithm and a novel DCM based attitude estimation algorithm which fits the GF-IMU configurations.
Publications
- "Constrained angular motion estimation in a gyro free IMU," IEEE Transactions on Aerospace and Electronic Systems vol. 47, 2011
E. Edwan, S. Knedlik, and O. Loffeld
- "Reduced DCM Based Attitude Estimation Using Low-cost IMU and Magnetometer Triad," in Proceedings of the 8th Workshop on Positioning, Navigation and Communication (WPNC'11) Dresden, Germany, 2011, pp. 1-6
E. Edwan, J. Zhang, J. Zhou, and O. Loffeld
- "A New Loosely Coupled DCM Based GPS/INS Integration Method," NAVIGATION, vol. 59, pp. 93-106, Summer 2012
E. Edwan, J. Zhou, J. Zhang, and O. Loffeld
- "Angular Motion and Attitude Estimation Using Fixed and Rotating Accelerometers Configuration," in Proceedings of the IEEE Position Location and Navigation Symposium (PLANS 2012), 2012
E. Edwan, J. Zhang, and O. Loffeld
- "Angular Motion Estimation Using Dynamic Models in a Gyro-Free IMU," Sensors Journal, vol. 12, 2012
E. Edwan, S. Knedlik, and O. Loffeld
- "Novel Approaches for Improved Performance of Inertial Sensors and Integrated Navigation Systems." PhD: University of Siegen, 2013
E. Edwan
