信号处理2026，Vol.42Issue(6)：797-809,13.DOI:10.12466/xhcl.2026.06.003
基于因子分解组稀疏正则化的无人机目标主体与微动部件回波信号分离
Separation of UAV Target Object and Micro-Motion Component Echo Signals Based on Factor Group-Sparse Regularization
摘要
Abstract
With the widespread adoption of unmanned aerial vehicle(UAV)technology,the threat posed by UAVs to low-altitude safety has become increasingly prominent.Effective radar detection and identification of UAVs has thus become critical.The multicomponent UAV echo signal is primarily composed of relatively stable main body echoes(from the airframe)and time-varying micro-motion component echoes(from rotors and propellers).The micro-Doppler effect generated by these moving components is a key feature in target identification.However,in inverse synthetic aperture radar imaging,it degrades the clarity of the main body image.Conversely,during feature extraction,the main body echoes obscure the micro-Doppler information.Therefore,achieving effective separation between the main body and moving component echoes is one of the core challenges in the effective detection and identification of UAV targets.A Hankel matrix low-rank sparse decomposition is proposed for the echo separation problem of the UAV main body and its micro-motion components.First,the subject and micro-motion signal separation problem is modeled as a low-rank sparse matrix decomposition problem.Second,to enhance the robustness of the algorithm,the factor group-sparse regularization method is employed to relax the rank function in the low-rank sparse decomposition model.This approach is then combined with the method of linearized alternating direction of multiples to solve the model,whereby the echo signals are separated from the narrowband radar target subject and micro-motion components.Finally,the effectiveness and robustness of the proposed method are verified through simulations and measured data processing.关键词
微多普勒/汉克尔矩阵/低秩稀疏矩阵分解/因子分解组稀疏正则化Key words
micro-Doppler/Hankel matrix/low-rank sparse decomposition/factor group-sparse regularization分类
信息技术与安全科学引用本文复制引用
王欢,李开明,宫志华,王玉强..基于因子分解组稀疏正则化的无人机目标主体与微动部件回波信号分离[J].信号处理,2026,42(6):797-809,13.基金项目
国家自然科学基金(62371468,62531020) The National Natural Science Foundation of China(62371468,62531020) （62371468,62531020）
