基于广义互相关与混沌序列预测集成算法的微瞬态电磁辐射信号检测方法  

Detection of Weak Transient Electromagnetic Signals Based on Generalized Cross Correlation and Chaotic Time Series Prediction Integrate Algorithm

作　　者：张悦[1] 刘尚合[1] 刘卫东[1] 

机构地区：[1]军械工程学院静电与电磁防护研究所,石家庄050003

出　　处：《电子与信息学报》2015年第11期2769-2775,共7页Journal of Electronics & Information Technology

基　　金：国家自然科学基金(61172035)~~

摘　　要：该文以研究微弱瞬态电磁辐射信号的探测手段为目的,提出广义互相关与混沌序列预测相集成的检测算法。基于双天线测试和互相关信息估计方法,将低信噪比条件下非周期微弱放电信号的检测问题转化为周期性时延参数的估计问题,同时降低了噪声水平。基于混沌预测方法对其估计结果进行预测,得到的误差均值即为目标信号检测结果。通过仿真与实验分析对该方法的可行性进行了检验,结果表明:在低信噪比条件下利用集成方法可有效抑制噪声干扰的影响,相对于传统互相关估计法或混沌预测法而言,对微弱瞬态电磁辐射信号具有较高的检测概率,且需要的脉冲积累次数较少,提高了检测效率,比较适合用于微弱电磁辐射源的远距离探测。To investigate the remote detecting approaches of weak transient electromagnetic signals, the detecting method based on the general cross correlation and chaotic time series prediction integrate algorithm is proposed. Based on the double antennas test and cross correlation information estimation, the signal detection of weak non-periodic discharge signal in low Signal to Noise Ratio （SNR） is transformed to the estimation of periodic time- delay parameters, which decreases the level of noise simultaneously. The results of estimation are predicted based on chaotic predicting method, and the mean value of predicting error is the detection results of target signal. The feasibility of the approach is analyzed by simulating and experimental method. The results show that in the low SNR, the integration method can effectively restrain the interference from noise. Compared with the traditional cross correlation method or chaotic prediction algorithm, the detecting probability is higher for weak transient electromagnetic signals. Furthermore, the pulse integrating algorithm needs the less accumulation times, and its detecting efficiency increases. Hence, the proposed integrate algorithm is suitable for remote detection of partial discharge source.

关 键 词：信号检测 瞬态电磁辐射 微弱信号 广义互相关 混沌 

分 类 号：TM937[电气工程—电力电子与电力传动]