Analysis On Water Demand Structure Change And Water Consumption Forecasting In Baotou

In recent years,with the quickening paces of changing from country to urban,urban watersupply is increasing day by day,as a result,the problem on water resources shortage isbecoming more and more stressful,so it is of great significance to study the evolution of urbanwater resources demand and analysis and forecast the urban water resources demandscientifically for making urban macroscopic economic development planning,water resourcesplanning,rational allocation of water resources,and taking the corresponding water-savingmeasures,to ensure the city economy develop sustainably.Based on the water demand data and some factors influencing the change of watersupply in2000~2010of Baotou City, Analysised the development trend of water demandstructure by using information entropy,meanwhile,using gray correlation analysis determinethe main factors influence the water change.Then established BP neural network model,GrayGM(1,1) model,Exponential Smoothing model,and Gray-exponential model to forecast thewater consumption in2009and2010of Baotou City,and compared their forecast accuracy,theresult shows:(1)The characteristics of water consumption structure of Baotou City is agriculturalwater demand accounted the dominant position,while the industrial and domestic waterconsumption occupied for a small proportion.In2000~2010,the total water consumptionpresented a tendency of slight rise,with an annual increase of473.76ten thousand m~3.Agricultural water consumption and its proportion decreased gradually,reduced from74517.37ten thousand m~3and75.31%in2000to69969ten thousand m~3and67.48%in2010.While industrial water consumption and its proportion increased gradually,increasedfrom17135ten thousand m~3and17.32%in2000to26157ten thousand m~3and25.23%in2010.Domestic water consumption increased slowly,with an annual increase of26.40tenthousand m~3,while its proportion was basically stable,maintain at around7.48%.In the11years,the information entropy of water consumption structure in Baotou City have a trend of increasing,but the amplification is not very notable,just increased from0.71in2000to0.80in2010.(2)Near11years,the gray correlation degree between total water consumption andpopulation,GDP,total industrial output value,built-up area greenbelt cover rate,cultivatedarea,and Industrial water recycling rate is higher,it shows that these factors are the mainfactors influence the variation of water demand of Baotou City.(3) BP neural network model,Gray GM(1,1) model,Exponential Smoothing model,andGray-exponential model can be used for the prediction of annul water consumption in Baotoucity,simulated the water consumption in2000~2008of Baotou with this four method,theMean Square Relative Error separately is0.42%,7.18%,9.55%and6.95%,so the simulationaccuracy of BP neural network is highest.Forecast the water consumption in2009and2010ofBaotou City with BP neural network model,the relative errors separately is-0.32%and2.23%.It shows that BP neural network model is the best model to forecast the waterconsumption of Baotou City.Water consumption which is influenced by many factors can be forecasted by the bestBP neural network model.