因无人机POS系统中GPS天线相位中心与航摄仪投影中心间总有一段空间距离，惯性测量装置（IMU）三轴陀螺坐标系与航摄仪像空间坐标系之间总存在角度偏差，为了从GPS天线相位中心的三维坐标以及IMU的姿态参数入手得到航摄仪投影中心的坐标和航摄像片的姿态角，利用光线束区域网联合平差法对测区进行空中三角测量。论文在系统分析国内外关于POS数据优化方法研究进展的基础上，以西和县测区为例，主要做了两个方面的研究： （1）从GPS天线相位中心、IMU姿态角及航摄仪投影中心三者之间的空间几何关系出发建立了POS系统摄站坐标、IMU姿态角、像点坐标三类不同的观测值方程，根据这些观测值方程进一步求解出既有摄影测量观测值又有非摄影测量观测值的联合平差数学模型，按最小二乘原理可得到POS数据的平差值。 （2）在此联合平差中，分别选用单位定权法和经验定权法以及验后方差分量估计中的赫尔默特法对观测值进行定权，基于Visual Studio 2010集成开发环境和.NET平台用C#语言编制了一套光线束区域网联合平差系统，并在测区进行实验验证，分析不同的定权方法对空三加密结果精度的影响规律、重点分析不同的定权方法对优化后的POS数据精度的影响情况。 结果表明，GPS天线安装时应尽量靠近航摄仪投影中心以减小定位误差，IMU陀螺坐标系的三轴应和航摄仪像空间辅助坐标系相应的三轴尽量一致，以确保像片的姿态角更准确，从而减少定向误差；平差中采用传统定权中的单位权定权和经验定权对最终的平差结果的影响并不是很大，只不过选用经验定权相比选用单位定权而言，其整个平差过程中的迭代次数少5次；平差中采用赫尔默特方差分量估计法定权，求解的地面点精度相比传统定权方法提高了不少，利用赫尔默特方差分量估计获取的POS数据平差值的精度比利用传统方法获取的精度高出将近一半；无论如何优化POS数据，其加密生成的地面点高程精度总小于平面精度。

There is always a spatial distance between the phase center of the GPS antenna and the projection center of the aerial camera in the POS system, and there is always an angle deviation between the coordinate system of the inertial measurement device (IMU) and the aerial camera. In order to obtain the coordinates of the projection center of the aerial projector and the attitude angle of the aerial image from the three-dimensional coordinates of the center of the GPS antenna in the POS data and the attitude parameters of the IMU, the aerial triangulation of the survey area is carried out by the beam leveling method. Based on the systematic analysis of the research of POS data optimization method at home and abroad, this paper takes the survey area of Xihe county as an example for study: (1) the geometric spatial relationship between the three of the projection center is established from the GPS antenna phase center, IMU attitude angle and aerial camera, so are the equations of the coordinate system of the POS system, the IMU attitude angle and the image coordinates. By these observation equations, Non-photogrammetric observations of the joint adjustment of the mathematical model, according to the least squares principle, can be obtained by POS data adjustment value. (2) For this joint adjustment, we use the unit weight method and the empirical weighting method and the Helmert in the posterior variance component estimation to determine the observation value. Based on Visual Studio10 integrated development environment and NET platform, a set of light beam area network joint adjustment system is compiled in C # language, and the experiment is carried out in the survey area. The influence law of different fixed weight method on the accuracy of empty three encryption results is analyzed. The experiment shows that the GPS antenna should be installed as close as possible to the astronomical projection center to reduce the positioning error. The three axes of the IMU gyro coordinate system should be consistent with the corresponding three axes of the camera-assisted coordinate system to ensure that the attitude angle is more accurate, so as to reduce the directional error; adjustment in the traditional fixed right of the unit rights and experience of the final decision on the final impact of the results is not very large, but the choice of experience compared to the use of units the number of iterations in the whole adjustment process is less. In the adjustment, the Helmut variance component is used to estimate the law, and the accuracy of the ground point is improved compared with that of the traditional method. The precision of the POS data obtained by the Helmert method is much higher than that obtained by the traditional method. In any case, the accuracy of the ground point is less than the plane accuracy.