The wavelet decomposition uses a pairs of filters iteratively to decompose the original series,It results in a hierarchy of new time series that are easier to model and predict.

• 其中,小波作为滤波部分对原始序列进行多尺度分解,产生更容易建模和预测的子序列,再把上述子序列作为神经网络的输入进行时域预报。