Study On The Combination Method Of Short-term Load Forecasting Considering The Accumulation Effect Of Temperature

Short-term load forecasting is a very important element in power system scheduling, which is to forecast the real time electricity consumption of the next week or a few days. Accurate forecast of short-term electrical load is very important to the power system security and economy. As the electrical industry of our country is made to market step by step, load forecasting becomes more important to power system planning, operation and control.This paper mainly studies load analysis, load factors in electrical load and the development of forecasting arithmetic. Firstly, the analysis of the daily type, weather type, temperature, daily rainfall, humidity, wind speed and other factors on the daily load curve are researched. Secondly, the year periodicity, month periodicity and daily periodicity are analyzed, so we can explore the new load forecasting methods. Thirdly, this paper presents a quantitative formula for the temperature accumulative effect and proposes a combinatorial forecast method based on wavelet analysis and Artificial Neural Network considering the accumulation effect of temperature. The load sequence is decomposed by wavelet transform, the approximate sequence and the detail sequence in different frequency are obtained. Then, the approximate sequence is forecasted by the suited BP which is joined the accumulated temperature coefficient and other characteristic parameters, and the detail sequence are forecasted by the suited BP directly. Finally, the forecasted results of the sub-sequences are reconstructed and considered as the final forecasted result.Result of calculation example show that the proposed model can improve the accuracy of forecasting efficiently.