基于最小二乘的全局信息融合方法在分布式POS中的应用

A global information fusion method based on least square applied to distributed POS

分布式POS是一种基于惯性/卫星组合技术的多节点高精度时空测量系统,是实现多任务高精度航空遥感载荷成像的关键装置。受阵风湍流、载机机动等外部扰动,柔性长机翼产生多模挠曲变形,导致柔性基线效应,降低分布式POS多传感器信息融合精度。针对大变形机翼柔性基线效应导致的多传感器信息融合难题,提出了一种基于最小二乘的全局信息融合方法。在分布式POS主子传递对准基础上,采用基于最小二乘的正交多项式拟合方法,对分布式POS所有主子IMU的运动信息进行全局信息融合,充分利用各子IMU对应节点间的空间相关性,得到分布式POS各节点全局估计信息,提高分布式POS测量精度。最后,通过分布式POS地面模拟仿真实验验证,所提出的方法与传递对准方法相比,x轴基线精度平均提高了1.893 mm,y轴基线精度平均提高了0.461 mm,z轴基线精度平均提高了2.183 mm,水平姿态测量精度平均提高0.023°。

The distributed POS is a multi-node high precision space-time measurement system based on SINS/GPS integrated navigation system, which is a key to high accurate imaging device for multitask airborne loading remote sensors. However, because of multi-source disturbances such as gusts and platform maneuvers, the flexible long wing generates multi-mode flexural deflection, which produces a flexible baseline effect decreasing the precision of distributed POS. In order to solve the problem of the multi-sensors information fusion caused by the flexible baseline effect, a global information fusion method based on least squares is proposed. Firstly, the airborne flexible baseline deformation model is established. Secondly, the least squares orthogonal polynomial fitting method is used to fit all the sub-IMUs motion information of the distributed POS. The global estimated information of each node in the distributed POS which further improve the accuracy of distributed POS is obtained by making full use of the spatial correlation between the corresponding nodes of each sub-IMU. Finally, the ground simulation experiment of distributed POS is carried out to validate the proposed method. The results show that the accuracy of lever-arm is improved by 1.893 mm, 0.461 mm and 2.183 mm on average at x, y and z axis, respectively. And the accuracy of roll is improved by 0.023 ° on average.