苯乙烯是国民经济发展中重要的基础石化产品,其价格影响因素复杂,波动较大。本文基于灰色预测和指数平滑预测的基本理论,提出了一种新的基于灰色GM(1,1)模型和三次指数平滑模型的线性加权平均的组合预测方法,并以2010年1月到9月的苯乙烯周平均价格作为历史数据,对2010年10月的4个周平均价格进行了组合预测。结果表明,预测误差大大减小,预测精度显著提高。 Styrene is an important basic petrochemical product in the development of national economy. Its price has complicated and fluctuating influence factors. Based on the basic theories of gray prediction and exponential smoothing prediction, this paper proposes a new combined linear prediction method based on gray GM (1,1) model and cubic exponential smoothing model. From January to September in 2010, Of the average weekly price of styrene as a historical data, the October 2010 4 week average price for the combined forecast. The results show that the prediction error is greatly reduced and the prediction accuracy is significantly improved.