基于小波神经网络的舰载雷达目标识别方法

Radar Target Recognition with Wavelet based Neural Network

介绍了用于雷达目标识别的小波神经网络的结构与算法。该方法基于子波变换进行特征提取和识别，在最小均方能量准则下，用共轭梯度算法求解子波函数线性组合的尺度和延时参数，以及神经网络的权值。结果表明，该方法的实现结构具有子波的非线性，可有效解决雷达目标的识别问题。

Proposes an algorithm and a structure of Neural Network for wavelet which are used in radar target recognition.This algorithm is based on wavelet transform.The dialation and shifts which are the parameters of wavelet function and the weights of Neural Network structure are calculated by the conjugate gradient algorithm in terms of the latest mean squared energy criterion.The experimental results show that this structure can represent the nonlinear feature and this algorithm can solve efficiently the problem of radar target recognition.