论文部分内容阅读
针对径流量时间序列非线性、非平稳性的特点,基于丹江口水库1933~2001年的入库水量资料,采用紧致型小波神经网络预测水文序列,将小波基与输入向量的内积进行加权计算和反复训练,发挥小波变换良好的时频局域化性质及神经网络的自学习功能,再通过1961~2001年降水量和入库水量的对比,分析了降水和径流的变化过程。结果表明,径流量有减少的趋势。
According to the non-linear and non-stationary characteristics of time series of runoff, based on the inflow data of Danjiangkou Reservoir from 1933 to 2001, the compact wavelet neural network is used to predict the hydrological sequence and the inner product of wavelet base and input vector is weighted And repeated training, to play a good time-frequency localization of wavelet transform and neural network self-learning function, and then through the 1961 ~ 2001 precipitation and the amount of water compared to analyze the process of precipitation and runoff. The results show that the runoff has a tendency to decrease.