基于均匀设计的神经网络结构优化设计方法

Method of NN Structure Optimization Design Based on Uniform Design

介绍了均匀设计法的原理,并采用均匀设计法,对径向基函数(RBF)神经网络的结构进行了优化设计,并将设计好的神经网络应用于13位巴克码的脉冲压缩检测,仿真结果表明:采用经过优化设计的神经网络进行脉冲压缩信号检测,有助于提高雷达的探测性能,其信号的主旁瓣比优于传统方法。

Introduces principle of uniform design on which a project of optimization design for RBF NN structure is presented, then the designed NN is applied to pulse compression detection of 13 digits Barker code. The simulation results indicate that the NN after optimization design gains good effect on solving the detection of pulse compression signal, it is benefit to raise the detection ability of radar, and the main lobe cluter rate is higher than traditional methods.