基于预报-校正法的城市用水量预测模型的研究

神经网络已成为提供预测与决策支持的有力工具,但传统的神经网络预测方法有一缺点就是对于峰值和谷值预测效果不佳,其预测值和实际值常有较大的误差.通过引入预报-校正法的思想,在传统的神经网络预测方法基础上另外构建一个独立的神经网络用于学习误差估计,加以对传统的神经网络预测值进行校正.实验结果表明,该方法能改进传统的神经网络预测的缺点,显著改善预测效果.此外,该方法在类似的基于神经网络的函数逼近等方面也有很好的改进效果.

Prediction model of city water supply based on prediction-revise

Neural network is a powerful tool for forecasting and decision-making, but its disadvantage is not good at forecasting peak value and vale value. A forecast model of city water supply based on neural network and prediction-revise is presented. The second neural network studies the prediction error of swatch in order to correct the prediction result of the first neural network. The experimental result show that this method is effective. In addition, this method is effective as well as in function approach.