摘要
自回归(Autoregressive,AR)模型是常用的平稳序列的拟合模型之一.不同于传统趋势的曲线拟合方法,文章利用经验模态分解方法(Empirical Mode Decomposition,EMD)得出序列的趋势,然后再进行建模.通过运城市某年的每日最高气温的数据,建模结果表明:与传统的建模方法比较,基于EMD的AR模型在拟合效果和预测效果上要更好.
The autoregressive model is one of the commonly used stationary series fitting models. Different from the traditional trend curve fitting method,It adopts EMD to get the trend of series and then model. By the data of daily highest temperature of Yuncheng city in 2005, the outcome of model shows that, compared with the traditional model, the AR model based on the EMD has better fitting effect and prediction effect.
出处
《太原师范学院学报（自然科学版）》
2013年第3期98-100,共3页
Journal of Taiyuan Normal University:Natural Science Edition
