Research on BP Neural Network Model for Water Demand Forecasting and Its Application

摘要

It is difficult to determine a proper neurons number of the mid-layer when using the BP neural network for water demand forecasting. Aiming at the problem, the BP neural network is presented in this paper for water demand forecasting. A suitable neurons number in the mid-layer is calculated based on the empirical formula method and trial and error method. A certain basin in China is taken as a case study. The results indicate that the mean relative error is 2.42%. The water consumption is 42.8 billion m~3 in 2015 and 43.6 billion m~3 in 2030 in the study area. The results are useful for water resources planning and management.