Torpedo electromagnetic fuze active interference recognition based on improved AlexNet

基于改进AlexNet的水中兵器电磁引信有源干扰识别方法

¹ School of Marine Science and Technology, Northwestern Polytechnical University, Xi'an 710072, China
² The 705 Research Institute, China Shipbuilding Industry Corporation, Xi'an 710075, China

Received: 24 October 2023

Abstract

In order to enhance the active interference signal recognition capability of the electromagnetic fuze of a torpedo in the marine electromagnetic environment, a method for recognizing the electromagnetic fuze's active interference based on the improved AlexNet is proposed. Interference signal models and marine electromagnetic wave channel models are established. The time-frequency plots of interference signals propagated through the marine electromagnetic wave channel are extracted into datasets. The trained network model is improved by optimizing the network structure, adding data labels and augmenting data. The recognition rate and complexity of the model under different jamming-to-noise ratio(JNR) conditions are provided. Simulation results demonstrate that the proposed method achieves a high interference recognition rate under the low JNR condition, accurately recognizes the electromagnetic fuze's active interference and exhibits low model complexity.

摘要

为了提升海洋电磁环境下水中兵器电磁引信有源干扰信号的识别能力, 提出了一种基于改进AlexNet的水中兵器电磁引信有源干扰识别方法。建立了干扰信号模型与海洋电磁波信道模型, 提取经海洋电磁波信道传播后的干扰信号时频图作为数据集, 通过优化网络结构、增加数据标签、数据增强等方法改进训练出的网络模型, 给出了不同干噪比(jamming-to-noise ratio, JNR)条件下的模型识别率及复杂度。仿真结果表明, 文中所提方法能在低JNR条件下达到较高的干扰识别率, 能够准确识别水中兵器电磁引信有源干扰, 同时具有较低的模型复杂度。