基于自适应指数平滑法的甲醇价格预测  

Methanol Price Prediction Based on Adaptive Exponential Smoothing Method

作　　者：赵天泽 胡学友[1] 张胜 ZHAO Tian-ze;HU Xue-you;ZHANG Sheng(School of Advanced Manufacturing Engineering,Hefei University,Hefei,230601,Anhui)

机构地区：[1]合肥学院先进制造工程学院,安徽合肥230601

出　　处：《蚌埠学院学报》2023年第2期113-117,共5页Journal of Bengbu University

基　　金：安徽省高校自然科学研究项目(KJ2021A079);合肥学院重点学科(2018xk03)。

摘　　要：传统的指数平滑预测法难以适应时间,在数据变化较大的情况下,无法根据数据做出实时的改变。提出一种随预测时间与实际值变化幅度动态改变的自适应指数平滑法,根据历史数据预测了未来的甲醇价格波动。预测实验表明,所提出的方法可以有效提高预测的精准度,优于原先单一的指数平滑法,具有更好的预测性能。The traditional exponential smoothing prediction method is difficult to adapt to the time,and the data changes greatly,it is impossible to make real-time changes according to the data.It presented an adaptive exponential smoothing method in this paper which dynamically changes with the forecast time and the actual value change amplitude,and predicts the future methanol price fluctuation according to the historical data.The prediction experiment showed that the proposed method can effectively improve the prediction accuracy,and has better prediction performance than the original single exponential smoothing method.

关 键 词：实时预测 动态指数平滑法 价格预测 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论] TP39[自动化与计算机技术—计算机科学与技术]