灰色神经网络最优权组合模型预测城市需水量

需水量预测是一个大量数据指标和影响因素共同作用的复杂系统。目前以单一的模型预测为主,而这种预测方法仅能体现该系统的局部。针对这一情况,利用灰色模型和改进BP神经网络,建立最优权组合模型预测城市需水量,使用Matlab进行实例计算,并与其他预测方法比较。结果表明,该模型有较高的预测精度,优于单个模型,预测效果更优于其他方法。

Urban Water Demand Forecasting by Combining Improved BP Neural Network and Grey Model with Optimum Weight

Prediction of water demand is a complex system affected by a mass of data and influencing factors together.Nowadays,most of the forecasting methods are single model ones.They reflect only part of the system.In view of this situation,a combined model is set up by use of grey model and improved BP neural network with optimum weight to forecast the urban water demand.It carries on the example computation by Matlab and compares the results with those by other prediction methods.It shows that the combined model has the high precision prediction.Its prediction results surpass those by the single model,even more surpass those by other ways.