Abstract: The classical econometrics models and forecasting methods based on point data have poor forecasting quality because those models might neglect the important interval information of the price movement.A new interval method based on interval computing and interval econometrics to predict crude oil prices is applied.Compared with the confidence interval forecasts based AR-GARCH model，the interval econometrics method yields more accuracy ratio and much less forecast error.This study validates the interval econometrics method，and uncovers the important application to economic issues.

Key words: interval computing, interval econometrics, interval forecasting, crude oil future’s prices

摘要： 经典的计量经济学建模与预测方法是基于点数据的，忽略了区间内价格波动的大量信息，因而预测效果欠佳。引入区间计算与区间计量方法，应用于国际原油期货价格的预测，研究结果表明：相对于经典AR-GARCH模型的置信区间预测结果，区间计量方法的预测结果具有更高的准确度与更小的预测误差。研究证实了区间计算与区间计量方法的优越性，并揭示了在经济领域的重要应用价值。

关键词: 区间计算, 区间计量, 区间预测, 原油期货价格