基于小波神经网络的城市用水量长期预测研究

Research on Long-term Prediction of Urban Water Consumption based on Wavelet Neural Network

针对城市年用水量时间序列的非线性和多时间尺度特性,应用小波分析方法对年用水量时间序列的趋势性、周期性和随机性进行了分析。在此基础上,建立了用于城市用水量长期预测的小波神经网络耦合模型。并利用此模型对昆明市年用水量进行预测,预测结果表明,该模型应用于城市用水量的长期预测具有较高的预测精度和良好的推广能力。

According to the nonlinear and the multi-time scale character of the annual water consumption time series of urban, the method of wavelet analysis was used to analyze the tendency, periodicity and random of annual water consumption time series. Based on this, a long-term predicting wavelet neural network coupling model of urban water consumption was built, and used to predict the annual water consumption of Kunming. The result showed that this wavelet neural network predicting model had good quality in terms of prediction precision and generalization.