中文摘要：

针对宽频馈源致冷测试实验中出现的某些异常现象和测试数据波动性较大的问题，仿真分析了可能原因，为后续测试提供参考。基本结论为，应确保系统增益与测试仪器线性范围的匹配，尽量减小输出功率的测试误差；应尽量增加带宽内测量频点的数目，以利于判别驻波和编辑与筛选测量数据；可依据Y因子的合理取值范围甄别测量粗差；将馈源向下对准液氮表面，在理论上更具自洽性。

英文摘要：

Simulation analysis was performed in order to reveal possible causes of some abnormal phenomena and obvious scatter in the data of equivalent noise temperature t of cooled broadband feed, and to provide some reference for sub- sequent tests. Preliminary conclusions are as followings. The system gain should match the linear response scope of the instrument to constrain the uncertainties of the output power measurements. To increase the number of frequency points within the bandwidth facilitates would be benefit to the identification of standing waves and the editing and selection of measurement data. The reasona- ble limitation in figure of the Y - factor could act as an indicator of outliers in the measurement. It is theoretically more consistent to let the feed upside down toward to the surface of liquid nitrogen.