是一种双隐层反向传播神经网络，可以得到很精确的幅值、频率、相位估计，包含特征值与特征向量的提取、训练样本以及最后的识别。- Is a two hidden layer back propagation neural network, You can get a very accurate amplitude, frequency, phase estimation, Contains the eigenvalue and eigenvector extraction, the training