基于深度学习的辐射场积分脉冲功率校准  

A Calibration Method of Radiation Field Integrated Pulse Power Based on Deep Learning Algorithm

作　　者：高启轩 彭博 叶娟 王淞宇 齐万泉 Gao Qixuan;Peng Bo;Ye Juan

机构地区：[1]北京无线电计量测试研究所

出　　处：《安全与电磁兼容》2025年第2期50-53,共4页Safety & EMC

摘　　要：脉冲大功率广泛应用于脉冲雷达、相控阵雷达、遥测等领域,其测量结果准确性直接影响电子设备在电磁环境适应性评估中的可靠性。辐射场积分脉冲功率测量方法是一种常用的脉冲大功率测量手段,但测量结果稳定性和一致性较差。文章基于深度学习算法建立了辐射场积分脉冲功率测量模型,通过深度神经网络修正辐射面的功率密度测量偏差,并进行了验证。结果表明,基于深度学习的脉冲大功率测量方法比传统辐射场积分脉冲功率测量方法的结果偏差减少了1.3 dB。研究成果可以拓展脉冲大功率校准领域的技术储备,进一步满足脉冲功率参数的量值溯源需求。Pulse high-power is widely used in pulse radar,phased-array radar,telemetry and other areas,and its calibration results directly affect the accuracy of electromagnetic environment monitoring and the reliability of electronic equipments adaptability evaluation in electromagnetic environment.Radiation field integrated power measurement method is commonly used in pulse high-power measurement,however,the measurement results are lack of stability and consistency.This letter establishes a radiation field integrated power measurement correction model,which is based on deep learning algorithm,and corrects power density measurement deviation of the radiation surface through deep neural networks.A verification is showed,and the experimental results show that pulse high-power measurement method based on deep learning algorithm reduces the measurement deviation by 1.3dB compared to traditional radiation field integration method.The research of this paper could expand technical reserve in pulse high-power calibration,so as to meet demands of pulse power parameter value traceability.

关 键 词：脉冲大功率 辐射场积分 深度学习 

分 类 号：TN9[电子电信—信息与通信工程]