Based On The Arima Model And Regression Analysis Of Electricity Consumption Forecast Method Research

Power consumption forecasting is the premise and foundation of the power system planning and economic operation. Through the analysis and research of historical data to find out the internal variation of power data and the correlation between power data and its influencing factors. Then make preliminary estimates of the demand for electricity. The accuracy of the results of power consumption forecasting is directly related to the power grid security and reliable power supply. At the same time it’s able to influence business decisions and enconomic benefits of the power grid enterprises.The analysis of power has the characteristics of uncertainty, conditionity, multivariant and timeliness. And this paper takes the short-term forecasting analysis. So choose the proper method to forecast and analysis the short-term power consumption. This paper:(1) It puts forward the two perspectives of analysis and forecasting ideas—longitudinal analysis and transverse analysis. longitudinal analysis is based on the autocorrelation properties between electricity data to establish ARIMA model and make predictive analysis; transverse analysis is based on the cross-correlation between the data of power consumption and its influencing factors to make predictive analysis, it combines the rough sets theory and regression analysis.(2) Research of short-term forecasting of power consumption based on ARIMA model. Elaborated the ARIMA modeling steps and its characteristics as a forecasting method. By the analysis of short-term forecasting of power consumption found that the precision of the forecasting results is very good as long as there is enough time series data.(3) Research and analysis of the regreesion method based on rough sets theory. The rough sets theory is introduced in the regression analysis. Using attribute reduction theory to solve the multicollinearity problem gives full play to the advantages of attribute reduction and regression analysis. This paper gives a detailed introduction to the method of steps, while makes a simple comparative with stepwise regression analysis.Paper emphatically analyze the regional electricity consumption forecasting from various angles, to improve the credibility of results. Using rough sets to data pretreatment, while the theory is introduced to the regression analysis to improve accuracy. The experiment results show that the prediction methods have good feasibility and accuracy.