| 李梦杰,行鸿彦,吴涵.基于IVMD-WPD-HH0-LSTM的海杂波小目标检测方法[J].电子测量与仪器学报,2026,40(1):156-168 |
| 基于IVMD-WPD-HH0-LSTM的海杂波小目标检测方法 |
| The sea clutter weak target detection method based onIVMD-WPD-HHO-LSTM |
| |
| DOI： |
| 中文关键词: 海杂波去噪 小波包分析 多阈值准则 HHO-LSTM 微弱目标检测 |
| 英文关键词:sea clutter denoising adaptive wavelet packet HHO-LSTM weak target detection |
| 基金项目:国家自然科学基金（62171228）、南通理工学院科技创新与服务地方团队项目(KCTD008)资助 |
|
|
| Author | Institution |
| Li Mengjie | 1.School of Electronics and Information Engineering, Nanjing University of Information Science & Technology, Nanjing
210044, China; 2.School of Electrical and Energy Engineering, Nantong Institute of Technology, Nantong 226001, China |
| Xing Hongyan | 1.School of Electronics and Information Engineering, Nanjing University of Information Science & Technology, Nanjing
210044, China; 2.School of Electrical and Energy Engineering, Nantong Institute of Technology, Nantong 226001, China |
| Wu Han | 1.School of Electronics and Information Engineering, Nanjing University of Information Science & Technology, Nanjing
210044, China; 2.School of Electrical and Energy Engineering, Nantong Institute of Technology, Nantong 226001, China |
|
| 摘要点击次数: 762 |
| 全文下载次数: 574 |
| 中文摘要: |
| 为提高海杂波背景下微弱目标的检测性能，提出了一种基于改进变分模态分解(improved variational mode decomposition, IVMD)-小波包多阈值分解(wavelet packet multi-threshold decomposition, WPD)与哈里斯鹰优化算法（Harris Hawks optimization, HHO）优化长短期记忆网络（long short-term memory, LSTM）相结合的检测方法。采用改进变分模态分解-小波包多阈值分解去噪技术，通过自适应粒子群（adaptive particle swarm optimization, APSO）优化确定变分模态分解最优参数，精确分解海杂波信号为多个本征模态函数(intrinsic mode functions, IMF)。针对噪声较强的高频IMF，设计多频段小波包分解与分层阈值去噪策略，有效抑制噪声并保留微弱目标特征。利用哈里斯鹰算法对长短期记忆网络模型的超参数进行优化，提升其在复杂海杂波背景下的非线性时序建模能力。通过相空间重构与去噪信号的结合，显著提高了目标检测的精度和抗干扰能力。实验使用加拿大MCMaster大学IPIX雷达实测数据，结果表明，该方法在高信噪比和低信噪比环境下均显著提高了检测精度，相比传统长短期记忆网络方法，检测能力至少提高了35%。 |
| 英文摘要: |
| To enhance the detection performance of weak targets against sea clutter, this paper proposes a hybrid detection method. This method integrates improved variational mode decomposition (IVMD) combined with wavelet packet multi-threshold decomposition (WPD), and a long short-term memory (LSTM) network optimized by the Harris Hawks optimization (HHO) algorithm. The IVMD, whose optimal parameters are determined adaptively by adaptive particle swarm optimization (APSO), is employed to precisely decompose the sea clutter signal into several intrinsic mode functions (IMFs). For the high-frequency IMFs containing strong noise, a multi-band wavelet packet decomposition and layered threshold denoising strategy is designed within the WPD framework to effectively suppress noise while preserving weak target characteristics. The Harris Hawks optimization algorithm is utilized to optimize the hyperparameters of the LSTM model, thereby enhancing its capability for nonlinear time-series modeling within the complex sea clutter environment. By combining phase space reconstruction with the denoised signals, the accuracy and anti-interference capability of target detection are significantly improved. Experiments using the real-world IPIX radar dataset from McMaster University, Canada, demonstrate that the proposed method markedly improves detection accuracy under both high and low signal-to-noise ratio conditions. Compared to traditional LSTM-based methods, the detection capability is improved by at least 35%. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
