基于神经网络的电磁环境模型实时计算技术研究

Research on Fast Electromagnetic Environment Modeling with Neural Net

电磁环境建模对研究电子战具有重要意义。基于纯电磁理论方法计算空间电磁态势所需计算量较大,难以快速得到计算结果,无法满足仿真系统的实时性要求。利用神经网络计算复杂地形表面的电磁态势,计算时间相对纯电磁理论算法减少一个数量级,计算误差不大于2.8%,能够有效满足仿真系统对实时计算电磁态势的需求。

Electromagnetic Environment Modeling significantly contributes to Electronic Warfare research. Pure electromagnetic methods require huge amount of calculating resources, while couldn＇t provide electromagnetic field data to simulation system in real-time. Neural Net is in- troduced to Electromagnetic Environment Modeling, calculation results show that Neural Net reduces calculating time significantly, while the calculating error is less than 2.8%.