论文部分内容阅读
从大量的雷电干扰中识别出核爆电磁脉冲信号是核爆电磁脉冲探测的关键技术。传统的识别方法利用傅里叶变换的方法在频域范围内选择六个特征值,联合对信号进行逐一筛选。本文利用小波变换良好的时频局部化特性对两类信号分别进行了研究,只需利用三个特征值就可达到以前的识别精度,取得了较好的识别效果。这有可能使电磁脉冲探测手段进入核监测领域。
Identifying nuclear explosive electromagnetic pulse signals from a large number of lightning interferences is the key technology of nuclear explosion electromagnetic pulse detection. The traditional method of identification uses the Fourier transform method to select six eigenvalues in the frequency domain and jointly select the signals one by one. In this paper, the good time-frequency localization of wavelet transform is used to study the two types of signals, respectively. Only three eigenvalues can be used to achieve the former recognition accuracy, and a good recognition effect is achieved. This will make it possible for electromagnetic pulse detection to enter the field of nuclear monitoring.