子孔径拼接方法在大口径光学元件检测中发挥着重要的作用,关于子孔径拼接精度研究也受到广泛重视。子孔径斜率数据可由哈特曼探测器测得,被测光学元件上每个子孔径上的斜率数据通过最小二乘法进行拼接。测量过程中,测量数据不可避免含有随机噪声,这将影响拼接参数(如倾斜)的不确定度。推导了误差传递公式及评价拼接精度的公式,并分别计算了并行拼接和串行拼接中任意子孔径上每一点的拼接误差。在0.06s的随机噪声下,拼接斜率的统计误差与理论误差之间的差别在10-9 rad量级。模拟实验结果证实了所提出的拼接精度公式的正确性,可以用来评价拼接精度,并从理论上给出了并行拼接误差小于串行拼接误差的原因。 The method of subaperture splicing plays an important role in the detection of large aperture optics, and the study on the precision of subaperture splicing has also been paid more and more attention. The sub-aperture slope data can be measured by a Hartman detector and the slope data for each subaperture on the optical element under test is spliced ​​by the least square method. During the measurement, the measurement data inevitably contains random noise, which will affect the uncertainty of the splicing parameters (such as tilt). The error transfer formula and the formula for evaluating the stitching accuracy are derived. The stitching errors of each point on any sub-aperture in parallel stitching and serial stitching are calculated respectively. At a random noise of 0.06s, the difference between the statistical error of the stitching slope and the theoretical error is in the order of 10-9 rad. The simulation results verify the correctness of the proposed splicing accuracy formula, which can be used to evaluate the splicing accuracy and give the reason why the parallel splicing error is less than the serial splicing error theoretically.