Analyzing And Researching Of Time Series For Coal's Price Forecasting

Coal is main fuel for enterprises of iron and steel, the price of coal is important factor for enterprises to decide how to purchase fuel. Basing on this background, forcasting for the price of coal is very necessary.This paper takes coal's price from 1990 to 2004 as sample series, and forecasts the price of coal by using three single forecasting models and combination forecasting model through analyzing the influential factors for the price of coal. The satisfactory result of forecasting shows that the combination forecasting model is best. Above all, the first chapter of the paper introduces not only the background and the meaning of researching but also internal and external situation of researching; The second chapter introduces the basic theories of time series at first, then analyzes four forecasting models, such as multiple linear regression model, ARMA( p,q) model, cubic exponential smoothing model, and combination forecasting model based on induced ordered weighted averaging operators; Using multiple linear regression model, ARMA(p,q) model, and cubic exponential smoothing model, this paper forecasts the tendency of the price of coal from the third chapter to the fifth chapter; In order to raise the accuracy of forecasting, the paper forecasts the sample series with combination forecasting model based on induced ordered weighted averaging operators in the sixth chapter, then compares the residual of forecasting with other three results. The result shows that the accuracy of combination forecasting model based on induced ordered weighted averaging operators is best. At the end, I make a summary for the paper, and put forward the prospect of the future researching.