The prediction of domestic water consumption is the key to the urban water supply planning,and the core of it is to improve the accuracy of prediction.Since the error of the traditional GM(1,1)model mainly comes from the background value and the initial value,two improvement methods are a-dopted:introducing the power function to improve the background value weight structure,making the improved model dominated by new data;introducing the power function to reduce the oscillation of the original data,enhancing the adaptability of the original sequence.The two models are applied to the prediction of domestic water consumption in Henan Province and compared with the traditional GM(1,1)model.The results show that all tests of the improved model satisfy the requirements,and the medium-and long-term water consumption can be predicted.The domestic water consumption of Henan Province in 2025 is predicted to be 48.31×108 m3.The improved GM(1,1)model with opti-mized original value has good prediction effect and high accuracy,and can provide reference for local water resources protection and management.
optimized original valueoptimized background valueimproved GM(1,1)modelwater consumption prediction
维普
万方数据
