论文部分内容阅读
油价预测是能源市场研究的一个重要领域.本文基于季节调整技术和周期性分析技术,提出了一个新的预测模型-季节-谐波模型,其思路是分解-组合.通过对五种油品(WTI原油、Brent原油、无铅汽油、柴油和取暖油)价格的实证分析,与其它五种方法(ARIMA模型、指数平滑、Winters方法、EGARCH模型和逐步自回归)相比,季节-谐波模型取得了最好的预测效果.
Oil price forecasting is an important field of energy market research.This paper proposes a new forecasting model-season-harmonic model based on seasonal adjustment technique and periodic analysis technique, and its train of thought is decomposition-combination.Through the analysis of five kinds of oil products WTI crude oil, Brent crude oil, unleaded gasoline, diesel oil and heating oil). Compared with the other five methods (ARIMA model, exponential smoothing, Winters method, EGARCH model and stepwise autoregressive), seasonal-harmonic model Made the best prediction.