电子学报 ›› 2022, Vol. 50 ›› Issue (3): 557-566.DOI: 10.12263/DZXB.20201131

• 学术论文 • 上一篇    下一篇

多通道盲电磁辐射现场分离方法研究

刘胜, 王帮民, 张兰勇   

  1. 哈尔滨工程大学智能科学与工程学院,黑龙江 哈尔滨 150001
  • 收稿日期:2020-10-13 修回日期:2021-12-26 出版日期:2022-03-25
    • 作者简介:
    • 刘 胜 男,1957年生,教授,研究方向为随机系统最优估计与控制、舰船航行与姿态控制、鲁棒控制系统理论与应用、电磁兼容.E-mail:sch.liu@163.com
      王帮民 男,1990年生,河南南阳人.博士研究生,研究方向电磁信号处理,自适应滤波技术.
      张兰勇 男,1983年生,副教授,研究方向为集群智能控制,随机信号处理,船舶运动姿态控制.

Research on Multi-Channel Blind Electromagnetic Radiation Field Separation Method

LIU Sheng, WANG Bang-min, ZHANG Lan-yong   

  1. College of Intelligent Systems Science and Engineering, Harbin Engineering University, Harbin, Heilongjiang 150001, China
  • Received:2020-10-13 Revised:2021-12-26 Online:2022-03-25 Published:2022-03-25

摘要:

针对船舶系统在复杂电磁环境下设备电磁辐射的多通道欠定盲源分离问题,提出了一种基于小波分解(Wavelet Decomposition, WD)、相关性筛选以及独立成分分析(Independent Component Analysis,ICA)的多通道盲电磁辐射现场同步分离方法(Wavelet Independent Component Analysis,WICA).该方法采用小波分解,构造出虚拟通道,然后通过相关性筛选有效分量,最终利用ICA得到设备电磁辐射源信号.此方法不仅解决了欠定盲源分离问题,而且通过相关性筛选提高了ICA的有效性.本文以互相关系数(Cross Correlation Coefficient, CRC)和相干性(coherence)作为实船试验数据的度量指标,通过与(Ensemble Empirical Mode Decomposition-Independent Component Analysis, EEMD-ICA)算法和常规ICA算法的比较,验证了WICA算法在实际应用中的有效性.此方法可推广至航天、国防领域等系统级平台复杂电磁环境的现场测试,具有较高的实用价值.

关键词: 系统电磁辐射测试, 独立成分分析, 欠定盲源分离, 小波变换, 电磁环境

Abstract:

Aiming at the problem of multi-channel undetermined blind source separation of electromagnetic radiation from ship equipment in complex electromagnetic environment, a method wavelet independent component analysis(WICA) of multi-channel blind electromagnetic radiation field synchronous separation based on wavelet decomposition(WD), correlation screening and independent component analysis(ICA) is proposed. In this method, wavelet decomposition is used to construct a virtual channel, then effective components are screened through correlation, and finally ICA is used to obtain the electromagnetic radiation source signal of the equipment. In this paper, the cross correlation coefficient(CRC) and coherence are used as measurement indexes of real ship test data, and the effectiveness of WICA algorithm in practical application is verified by comparing ensemble empirical mode decomposition-independent component analysis(EEMD-ICA) algorithm. This method can be extended to the field test of complex electromagnetic environment of system-level platforms in the fields of aerospace and national defense, and has high practical value.

Key words: system electromagnetic radiation testing, independent component analysis, undetermined blind source separation, wavelet transform, electromagnetic environment

中图分类号: