基于变分模态分解的侵彻过载信号特征提取  

Feature Extraction of Penetration Overload Signal Based on Variational Mode Decomposition

作　　者：张晨阳 张亚[1] 李培英[2] 李世中[1] 赵海峰[3] ZHANG Chenyang;ZHANG Ya;LI Peiying;LI Shizhong;ZHAO Haifeng(School of Mechanical Engineering,North University of China,Taiyuan 030051,China;Handan College,Handan 056005,China;School of Mechanical Engineering,Nanjing Institute of Information Technology,Nanjing 210032,China)

机构地区：[1]中北大学机电工程学院,山西太原030051 [2]邯郸学院,河北邯郸056005 [3]南京信息职业技术学院机电学院,江苏南京210023

出　　处：《探测与控制学报》2021年第3期16-21,共6页Journal of Detection & Control

摘　　要：针对传统侵彻过载信号处理方法存在滤波效果不佳、模态混叠、端点效应、自适应性差的问题,提出基于变分模态分解的侵彻过载信号特征提取方法。该方法将侵彻过载信号的特征提取过程转移到变分框架内进行处理,通过寻找变分模型的最优解获取本征模态函数,能够自适应地实现信号的频域划分和各分量的有效分离,并有效地提取出侵彻过载信号的数据统计特性。实验验证结果表明,与经验模态分解相比,变分模态分解的特征提取效果更佳、信噪比更高,重构信号的积分结果更好地反映了弹体的实际侵彻深度,是一种用于侵彻实验数据事后处理的可行的新方法。Aiming at the problems of poor filtering effect,modal aliasing,end effect,and poor adaptability existing in the traditional processing methods of penetration overload signal,a characteristic extraction method of penetration overload signal based on variational modal decomposition was proposed.This method transfered the feature extraction process of the intrusion overload signal to the variational framework for processing,and obtained the eigenmode function by finding the optimal solution of the variational model,which could adaptively realize the frequency domain division of the signal and the analysis of each component effective separation,and effectively extracted the statistical characteristics of the data penetrating the overload signal.Experimental verification results showed that,compared with empirical mode decomposition,the feature extraction effect of variational mode decomposition was better and the signal-to-noise ratio was higher.The integration result of the reconstructed signal better reflected the actual penetration depth of the projectile.

关 键 词：侵彻过载信号 变分模态分解 重构信号 信噪比 侵彻深度 

分 类 号：TN911.7[电子电信—通信与信息系统]