Study On Monthly Load Forecasting Model Of Guangdong Province

Power system load forecasting refers to the process that counts the data in a certainperiod of time with the original survey data as the basis, starting from the history andcurrent situation of electricity use, considering the operation characteristics of systems,such as the nature of power supply, load, man-made with the capacity decision and natural,social conditions as well as the local economy development conditions, studying or using aset of practical system to deal with the past and the future load`s mathematical methodwhich under a certain accuracy sense, then determine the load value at a particular moment.Load forecasting provides not only the basis for drafting production planning anddevelopment for the production of electric power department and managementdepartment,but also future electricity data index of supply area. With the development ofnew theory and new technology, both at home and abroad for electric load forecastingproblem is still in the research, but when compared to the prediction of short-termload, research of medium and long-term load forecasting is relatively insufficient, thedifficulty also correspondingly larger.As the economic strong province,Guangdong`s total economic output ranks first ofChina in the past20years.The whole society electricity consumption is also growingrapidly, its GDP varying situation and the total electricity consumption tendency are insubstantial agreement, the impact of economic change for long-term load change is moreobvious, the economic factor should be reflected in the load forecasting. At present, themonthly load forecasting methods which are seen and used are widely divided into twocategories: one is the pure form prediction method that according to the annualdevelopment sequence monthly quantity;another kind is the prediction method that madeup of the sequence of similar monthly amount. Both of the two categories are analyzedbased on historical data, getting the monthly consumption growth of pure law of history forpredicting,these methods are very accurate during the economic stages which growsmoothly.However, when encounter the situation such as in2008or in2010when theeconomic growth and the consumption growth rate changes greatly, the two methods areprone to have great error.This essay proposes that bring the economic factors into the monthly loadforecasting. Studying the relationship between Guangdong province power growth andeconomic growth by using the theory of Econometrics, setting a time series stationarity test,Granger causality test and OLS estimates model, quantifying the influence of economic growth on power growth. Next to establish the multiple linear regression predictionmodelunder which under the economic impact to predict.Finally solving the specialproblems in January and February of load forecasting via the shift correction method andgetting Guangdong power monthly load forecasting model under the macroeconomicoperation.Finally, a simulation example had set and the model is applied to the power system ofGuangdong monthly load forecasting process,and make comparison with the actual load.The results show that, bringing the economic factors into the model are to some extentcould ensure the prediction accuracy, as well as aquire a better performance, which has acertain guiding significance for the monthly load forecasting.