基于MI-EMD的激光引信回波信号去噪方法研究  

Research on denoising method for laser fuze echo signal based on MI-EMD

作　　者：何海军 胡鹏飞 田博 苏宏[1,2] 李林豪 李铁[1,2] HE Haijun;HU Pengfei;TIAN Bo;SU Hong;LI Linhao;LI Tie(Science and Technology on Electromechanical Dynamic Control Laboratory,Xi’an 710065,China;Xi’an Institute of Electromechanical Information Technology,Xi’an 710065,China)

机构地区：[1]机电动态控制重点实验室,西安710065 [2]西安机电信息技术研究所,西安710065

出　　处：《兵器装备工程学报》2024年第5期285-290,共6页Journal of Ordnance Equipment Engineering

摘　　要：为解决激光引信回波信号容易被噪声污染、信噪比低的问题,提出一种基于互信息经验模态分解的激光引信回波信号去噪方法。该方法融合了互信息相关性和经验模态分解自适应性的特点,对基于雷达原理建立的激光引信回波信号进行经验模态分解,通过互信息及相关阈值区分噪声模态和信号模态,提取出分解信号中的有用信号分量,并将其相关模态进行重构实现噪声的有效去除。实验结果表明:该方法处理后的信噪比提高到了18.8854 dB,均方根误差减小到2.81×10^(-6),且去噪后信号曲线的平滑度较高。该方法能有效滤除激光引信回波信号中的噪声,很好地还原激光引信回波原始信号,保证信号的完整性,为后续激光引信在噪声条件下的精确定距奠定了坚实的基础。In order to solve the problem that laser fuze echo signal is easy to be polluted by noise and the signal-to-noise ratio is low,a method of laser fuze echo signal denoising based on mutual information empirical mode decomposition is proposed.This method integrates the characteristics of mutual information correlation and empirical modal decomposition adaptivity,and performs empirical modal decomposition of laser fuze echo signals built based on radar principle.Then the noise modes and signal modes are distinguished by mutual information and correlation threshold,the useful signal components in the decomposed signals are extracted,and their related modes are reconstructed to realize the effective remove of noise.The experimental results show that the signal-to-noise ratio of this method is improved to 18.8854 dB,the root-mean-square error reduced to 2.81×10^(-6),and the smoothness of the signal curve after denoising is relatively high.This method can effectively remove noise from the laser fuze echo signal,restore the original signal of the laser fuze echo,and ensure the integrity of the signal.This lays a solid foundation for the subsequent accurate distance calculation of laser fuze under noise conditions.

关 键 词：激光引信 回波信号 互信息 经验模态分解 去噪 

分 类 号：TN223[电子电信—物理电子学]