一种新型的雷达吸波结构快速设计方法

A Novel and Fast Method to Design Radar Absorbing Structure

该文提出了一种新型的基于神经网络和智能优化算法的雷达吸波结构快速设计方法,提高了雷达吸波结构设计的速度和效率。利用神经网络计算了吸波结构几何尺寸与频率响应之间的非线性关系,神经网络具有逼近精度高、泛化效果好的特点。选择粒子群算法作为优化器,根据优化目标获得最佳的几何尺寸。将神经网络与优化技术相结合,可显著降低吸波结构设计的复杂性。通过比较神经网络和仿真在相同优化参数下的输出结果,验证了所提方法的准确性。结果表明,该方法能快速有效地设计出符合要求的吸波结构。

This paper proposes a novel and fast method based on neural network and optimization algorithm to improve the speed and efficiency in designing Radar Absorbing Structure(RAS). The nonlinear relationship between geometric dimensions and frequency response of RAS was computed by neural network, which has the characteristic of accurate approximation and good generalization. The PSO algorithm was chosen as the optimizer to obtain the best geometric dimensions based on desired requirements. Through the combination of the neural network and optimization technique, the complexity of designing absorbing structure would be reduced significantly. The accuracy of the proposed method was verified through the comparisons between the outputs of neural network and simulation under the same optimized parameters. The results show that this method could design the absorbing structure effectively and quickly to meet the requirements.