基于HHT与LSSVM的远区核爆电磁脉冲识别

为实现远区核爆电磁脉冲（NEMP）和闪电电磁脉冲（LEMP）的有效识别，提出一种基于希尔伯特黄变换（HHT）和最小二乘支持向量机（LSSVM）的识别算法。采用希尔伯特黄变换对远区NEMP和LEMP进行分析，利用两种信号的Hilbert谱在不同频带上分布的差异性，选择谱图中两个区域的能量占比作为信号的特征，选择LSSVM作为分类器进行分类识别。实验结果表明，采用能量占比特征可有效识别NEMP和LEMP，且综合识别率可达到98.59%。

Recognition of far-region nuclear electromagnetic pulse based on Hilbert-Huang transform and least square support vector machine

To realize effective recognition of far-region nuclear electromagnetic pulse and lightning electromagnetic pulse, a recognition algorithm based on Hilbert-Huang transform (HHT) and least squares support vector machine was proposed. According to the difference in the distribution of the Hilbert spectrum resulted by applying HHT to NEMP and LEMP, the energy ratio of two different normalized frequency bands was selected as features of signals. In addition, least square support vector machine (LSSVM) was chosen as the classifier. A test using measured data was performed to verify the algorithm, which can effectively identify NEMP and LEMP data through the energy ratio in Hilbert spectrum with 98.59% recognition rate.