Radio astronomical signals are very weak, so the electromagnetic environment is very important for radio astronomical observations. Usually, electromagnetic interference can be reduced by means of favorable terrain, establishing radio quiet area, electromagnetic shielding and protection, etc. However, there are still some RFIs (Radio Frequency Interferences) that are difficult to block out. This paper proposed an RFI mitigation method with the adaptive filter, which can be used to extract astronomical signals in complex noise environment. It adopts adaptive transverse filter and the least mean square (LMS) error algorithm by changing the order and step size to optimize the filtering effect. The evaluation criteria are systematic errors and convergent performance. The simulation results show that the filter can effectively extract the signal while ensuring the convergence of the algorithm. In order to test the effectiveness of the algorithm, the observation data recorded by Nanshan 26 m radio telescope of Xinjiang Astronomical Observatory and Parkes 64 m radio telescope are selected, and the designed filter is used to test different measured data to verify the effectiveness of the filter. The agreement between theoretical analysis and experimental results shows that this method can effectively eliminate the interference signals in the astronomical observation and has certain practicabilities.