强噪声下非平稳时延目标信号提取

为实现强噪声背景下时延目标信号自提取，设计了相关检测模型，提出了基于目标数据段自截取和变分模态分解（Variational Mode Decomposition，VMD）算法的目标信号自适应定位和提取算法。首先,根据目标信号频域参数构建基准信号；然后,利用互相关检测技术并以峰值梯度为特征值确定目标信号时延参数；最后,利用预设参数自适应初始化的VMD算法对截取的数据段进行处理，并对分解的各分量进行自动筛选实现目标信号有效提取。不同算法的对比实验表明，自截取+VMD算法可有效实现强噪声下时延信号的定位提取，实时性基本满足实际应用需求。

Adaptive extraction of non-stationary delay target signal under strong noise

In order to realize adaptive extraction of delay target signal in strong noise background,a coherent detection model is designed and an adaptive target signal segment localization and extraction algorithm based on self-interception and variational mode decomposition(VMD) algorithm is developed.Firstly,the reference signal is constructed based on the frequency domain parameters of the target signal.Then,the target signal delay parameters are determined by the cross-correlation detection technique and taking the peak gradient as the characteristic parameter.Finally,VMD algorithm with adaptive preset parameters is used to process the extracted data segment,and decomposition components can be screened automatically to realize effective extraction of target signal.The comparative experiments of different algorithms show that the VMD algorithm after self-interception can effectively realize the location and extraction of delay signal under strong noise.The real-time performance of the algorithm basically meets the requirements of practical application.