Research On Forecasting System Of Thermoelectric Enterprises Fuel Consumption

Electricity is important foundation supporting the social and economic development, itplays an important role. Electric power as the double energy with usability, security,environmental protection, high efficiency, it is widely used in our country at present. Theenergy demand for electricity has increased with steady economic growth year by year. In theperiod for a long time, if we without the development and supporting of electric power, socialand economic development will be unable to move, the people’s living standards are notsatisfactory.Thermoelectric generation occupies the largest proportion in Chinese power structure,and the power structure will exist for a long time, the amount of thermoelectric generationaccounted will be kept around at70%. In order to improve the production efficiency andmanagement level, thermoelectric enterprises are trying to build and improve the informationconstruction of enterprises. In this process, many enterprises proposed to predict fuelconsumption demand from practical point of view. These enterprises hope to accuratelypredict the future number amount coal based on one month of daily fuel, to reduce theamount of coal storage, thereby reducing the coal enterprise loss in this part.Based on above requirements, this paper proposes the prediction method of thethermoelectric enterprises fuel amount based on ADF unit root test. The prediction methodcalculate and determine the amount of coal daily prediction calculation formula by the dailyamount of coal for three years based on the ADF unit root test, and predicting the daily coalamount of the next month. The basic process to determine the prediction formula based onADF unit root test: The first step to determine whether the time series of daily coal amountaccord simple exponential, the second step to determine whether the time series of daily coalamount accord the exponential correction, the third step to determine whether the time seriesof daily coal amount accord double exponential curve model, the fourth step to determinewhether the time series of daily coal amount accord two type, the fifth step is to determinewhether the time series of daily coal amount containing drift, trend stationery, the sixth step isto determine whether the time series of daily coal amount containing steady drift characteristics, the seventh step to determine whether the time series of daily coal amount isstable, each step of the judgment result can pointed out the time series of daily coal amountaccorded model and the required data processing method, or points out the the next stepjudgment process. This system not only realizes the test process, at the same time it realizesthe self-regression, exponential smoothing forecasting model. The fuel amount forecastingsystem of thermoelectric enterprises is used as an independent system, at the same time, thesystem also can be embedded into the enterprise production management information system(MIS) as the forecasting module provides forestation function as dosage of MIS fuel system.Based on the realization of this prediction method about forecasting system ofthermoelectric enterprises, can improve budget accuracy of the coal amount in a certain extent,so as to reduce the coal storage of electric power enterprises, and reduce the losses ofenterprises ultimately.