Prediction of Urban Hourly Water Consumption in Zhengzhou City

For supplying optimal scheduling of Zhengzhou city with water consumption data, this paper builds three types of forecasting model according to moving arithmetic mean method, regression analysis method and BP neural network. As a result, forecasting result is obtained by water supply data and meteorological data. The study shows that three different methods all can meet the need of urban water supply project. Regression analysis and BP neural network can obtain better foresting result and can gracefully satisfy the request of urban water supply scheduling.