针对射频干扰与回波信息在幅度和频域分布上的不同特性,提出了一种改进的经验模态分解(EMD)算法.该方法首先对包含回波信息和射频干扰的原始信号乘以给定的高频信号,将其搬移至高频带,然后通过EMD将回波信息和射频干扰分解到不同的本征模态函数(IMF),在每个IMF中对高出阈值的干扰部分置零,最后将处理后的IMF相加重构信号,并搬移回原频带.达到在抑制射频干扰的同时,无损干扰位置处的回波信息.该方法处理速度快捷,满足雷达实时处理要求.实测数据验证表明:提出的算法改进了原始EMD算法的缺点,并且在不损失回波信息的基础上最大程度地抑制了射频干扰. Aiming at the different characteristics of amplitude and frequency domain distribution of radio frequency interference and echo information, an improved empirical mode decomposition (EMD) algorithm is proposed, which first multiplies the original signal containing echo information and radio frequency interference by The high frequency signal is then shifted to a high frequency band, and then the EMD is used to decompose the echo information and radio frequency interference into different intrinsic mode functions (IMFs). The interference part above the threshold is set to zero in each IMF, Finally, the processed IMFs are added to reconstruct the signal and moved back to the original frequency band to reach the echo information at the position of non-destructive interference while suppressing the radio frequency interference.The method is fast in processing speed and meets the real-time radar processing requirements. It shows that the proposed algorithm improves the shortcomings of the original EMD algorithm and restrain the radio frequency interference to the maximum extent without loss of echo information.