即将开始建设的平方公里阵望远镜SKA将是全世界最大的射电望远镜。对SKA观测获得的可见度函数进行射频干扰去除或标注处理将是保证最终成像质量的重要环节。现有标注方法一般基于信号强度趋势变化分析，存在无法支持不同观测目标、准确率低、性能较差等问题，不能满足SKA数据处理的需要。本项目拟基于SKA仿真观测数据和先导阵实际数据，结合SKA数据处理的性能要求，重点研究并力争解决SKA可见度函数标注中存在的问题，具体研究利用仿真方法建立SKA仪器状态与可见度函数标准集，基于时序的多参数组合自动标记方法，基于机器学习的针对目标异常状况的自动标记方法，以及利用天空模型与脏图检验标注结果，对采用的可见度函数标注算法的可用性进行对比分析。项目研究成果可应用于SKA的数据处理，提高科学数据质量，同时也可为国内区域科学数据中心的数据处理提供支撑，推动后续科学研究工作的开展。

The Square Kilometre Array (SKA), which is about to build in near future, would be the largest radio telescope around the world. Due to the activities of human beings, it is necessary to mitigate Radio Frequency Interference (RFI) or to flag the visibility function so as to guarantee the imaging quality. However, current techniques on flagging have a series of drawbacks such as low accuracy, low performance and no supporting different observational objects, which is hard to meet the requirements of SKA data processing. In this study, considering the SKA data processing, we will focus on the challenges of visibility function flagging. A series of researches will be implemented including the construction of standard data sets of normal and abnormal apparatus status, multi-parameter combined flagging techniques based on time series, automatic flagging techniques based on machine learning, and result assessment technique based on sky-model and dirty map. The achievements of this study would solve the problem of RFI mitigation and contribute to the development of SKA telescope. Meanwhile, it is quite important for the construction of SKA Regional Center in the near future.