Research On The Prediction Of Water Consumption Based On The Self-organization Data Mining

In this paper, using GMDH method to do some researches of predicition and warning of the regional water consumption regarded as the research object of this paper.1、The research of index system for impact factors of regional water consumption.In the principle of macroscopic, this paper selects 14 indicators such as population, rainfall to build index system for impact factors, hailing from six aspects of influence factors of water consumption including population, resource, economy, science and thehnology, energy, management.2、The cluster analysis of regional economic development in China. This paper selects 19 macroeconomic development indicators such as per capita GDP、local financial income to build system clustering method, and the 31 provinces (municipalities) of China are divided into four categories of economic region.3、The forecast studies of regional water consumption.This paper chooses four regions, including Beijing, Guangdong, Henan and Gansu, and uses GMDH method by using programming of R.The paper designs the block diagram of structure of the warning of the regional water consumption, builds four forecast models of water consumption. The paper obtains the predicted values of each factor included in the above four prediction models and further obtains the predictive values of water consumption in 2014 and 2015.4、The warning analysis of regional water consumption.Basing on the water consumption in 2014 and 2015 and combining with water warning program, the paper indicates alarming degree of four regions, including Beijing, Guangdong, Henan and Gansu, and shows the GMDH method is an effective method for the early warning of water consumption.