Research on Electricity Demand Forecasting Based on Improved Grey Prediction Model

Simulation precision of conventional grey prediction models is poor when modeling sequence has the feature of oscillation. Actually, the smoother the sequence is, the higher the simulation precision is. With the purpose of perfecting the smoothness of oscillation sequence and improving the simulation precision of grey models, this paper researches a smoothing algorithm which can compress the amplitude of oscillation sequence, and by this algorithm, deduces a novel grey prediction model based on oscillation sequence, that is 〖AKx＾〗(t)=Fβt-31-（-1）tE-T. Finally, we employ the new model to forecast the electricity demand of a city in western China, and compare the simulation precision with other grey models(the simulation error of the new model is 7%,others are all 12%), the results shows the new model has the best simulation effect. Research findings in this paper have an important significance to enrich and perfect grey system theory and construct a more reasonable electricity demand forecasting model.