针对强电磁脉冲孔耦合瞬态响应的极点参数提取问题，将传统的矩阵束算法进行改进。通过对含有噪声的瞬态响应数据进行互相关处理，能有效提高低信噪比时极点的提取精度。再将数据经离散化后构造Hankle矩阵，结合奇异值分解，求解其广义特征值，即可提取信号的极点参数信息。最后通过仿真分析以及对波形的重构，验证改进算法具有抗噪能力强、拟合度高的特点。

To resolve transient response of the pole coupling parameter extraction problems from strong electromagnetic pulse, the traditional matrix pencil algorithm has been improved. Preprocessing the noisy transient response with inter-related processing method can effectively improve the extraction accuracy under low signal-to-noise ratio(SNR). The data are constructed as Hankle matrix after discretization, combined with singular value decomposition, for solving the generalized eigenvalue. Then, the signal poles parameter information can be extracted. It is verified that the improved algorithm has strong anti-noise ability and high fit through simulation analysis and reconstruction of the waveform.