Adaptive Unscented Two-Filter Smoother Applied to Transfer Alignment for ADPOS

作者：Lu, Zhaoxing; Li, Jianli^*; Fang, Jiancheng; Wang, Shicheng; Zou, Siyuan

来源：IEEE Sensors Journal, 2018, 18(8): 3410-3418.

DOI：10.1109/JSEN.2018.2799211

• Distributed position and orientation system
• transfer alignment
• flexible deformation
• mechanics modeling
• data fitting

The array antenna SAR or integrated multi-task imaging sensors are the most attractive development directions of aerial survey and remote sensing system, and urgently demand an airborne distributed position and orientation system (ADPOS), which depends on transfer alignment to accurately measure multi-node motion information. In this paper, an adaptive unscented two-filter smoother (AUTFS) is proposed for the offline transfer alignment of ADPOS, which deals with the problem of time-varying measurement noise covariance. The proposed smoother adopts a forward-backward approach, which includes an adaptive unscented Kalman filter tuned by fading factors in the forward direction and a backward filter using the weighted statistical linear regression formulation. A semi-physical simulation based on a flight experiment with ADPOS shows that the proposed AUTFS can achieve higher accuracy than the standard unscented two-filter smoother.

• 出版日期2018-4-15
• 单位北京航空航天大学