Cities Need To Forecast Water System Development

Urban water demand forecasting can be assorted into annual water demand forecasting and hourly water demand forecasting, daily water demand forecasting. They are efficient means of programming and managing of urban water resource, and they are important portion of optimizing dispatching management of water supply system.On the basis of analyzing historical water consumption in Shenzhen, hourly water demand, daily water demand and annual water demand are studied using non-linear regression model, time series model, artificial neural network, gray model and compounding model, etc. By anglicizing merits and demerits of every model in different forecasts, time series model is appropriate to hourly water demand forecast; compound forecasting model of time series and regress analysis is appropriate to daily water demand forecast; gray model and regress analysis model is appropriate to annual water demand forecast. The principle for urban water demand forecasting model is pointed out by comparing each model.In order to improve urban water demand forecasting precision, classified-forecasting method is presented. With its application in daily water demand forecast, daily water demand forecast is separated into domestic water consumption, industrial water consumption, commercial water consumption and common water consumption. It proved that the forecasting precision has been more accurate in a certain extent.On the basis of urban water demand forecasting methods' studying, and ActiveX controls integrated, urban water demand forecasting system for Shenzhen is developed by using Visual Basic, SQL Server and MATLAB as the developing stage. The system, which offers a simple and efficacious way to develop software, is effective in timely optimal control of water supply system, and the system is worth referring to while developing other optimal dispatching software such as water supply system.