在分析两路反馈外腔结构自混合干涉（SMI）模型的基础上，对两路自混合干涉信号的处理方法进行了理论分析和仿真验证。在弱反馈水平下，利用希尔伯特变换原理实现了两路合成信号相位的重构，并对合成信号相位分解得到了两路外腔反射体位移的波形，最大误差小于1.5 rad。在对含噪的两路自混合干涉信号进行预处理时，采用小波软阈值和集合经验模态分解（EEMD）相结合的算法，达到了良好的去噪效果。

Based on the analysis of the theory model of self-mixing interference (SMI) with two feedback external cavity structure, the processing method of self-mixing interference signal is analyzed and simulated. Under the weak feedback level, the reconstruction of synthetic signal phase and the shape reconstruction of two channel signals of external cavity are realized by using the Hilbert transform, and the maximum error is less than 1.5 rad. The self-mixing interference signal with noise is preprocessed by the wavelet soft threshold and ensemble empirical mode decomposition (EEMD) combining algorithm, which achieves a good effect.