射电望远镜指向误差提取及其容错修正分析
The Analysis of Pointing Error Extraction and Its Tolerant Calibration for a Large Radio Telescope

作为评估射电望远镜性能的重要指标之一，天线指向精度直接影响着射电望远镜的跟踪观测性能。本文采用十字扫描法对适合观测任务要求的目标射电源沿天线各方向指向进行测量，并采用误差提取算法提取天线指向误差。随后利用最小二乘法拟合得到指向误差修正模型系数，并提出一种基于残差统计检验对误差观测序列进行核验剔除离群值的容错修正方法。为检验指向误差修正模型的有效性，基于统计假设检验理论，利用模型拟合的F检验以及模型分量的t检验方法进行了显著性检验。观测实验验证结果表明：有效的指向误差与离群值的提取能够提高模型误差修正的精度，从33.46\"改进到26.39\"。显著性检验方法可用于复核误差模型以及模型分量的有效性。
As one of the most important characteristics for a radio telescope, the pointing accuracy of the antenna directly affects the observation performance of the radio telescope. The pointing measurements are collected via cross-scan method in the antenna’s hemispherical sky area, and the pointing error is extracted based on the Gaussian model. In this thesis, the error sources of the antenna pointing are studied and then analyzed the adopted correction methods. The coefficient values of the correction model are obtained by least squares method and a fault-tolerant correction method based on residual statistical test is proposed to verify the error observation sequence and eliminate outliers. Based on the theory of statistical hypothesis test, F-test algorithm and t-test algorithm are built to test the fitness of correction model and diagnose the significance of the model components. Several experiments with measured data demonstrate that the modified parameter model for the pointing corrections from 33.46\" to 26.39\". The hypothesis test can be used to identify the validity of error model and model components.