论文部分内容阅读
由于影响稀土产品价格的因素众多,各因素之间保持着错综复杂的联系,并且稀土产品月度价格数据具有高度的非平稳性、非线性和噪声的特性,增加了稀土产品价格预测难度。为此,运用时间序列预测法,以稀土Nd_2O_3,Dy_2O_3月度价格为例,建立非平稳时间序列ARIMA(1,1,2)模型,来描述并预测稀土产品价格的动态变化,得到了2006年1月~2015年12月的Nd_2O_3,Dy_2O_3价格预测值,并使用真实观测值与预测值进行预测拟合精度分析,结果表明,该模型拟合精度较高,适合于中短期模拟预测Nd_2O_3,Dy_2O_3产品价格。由于中国主要稀土产品价格的波动周期具有一定的相似性,表明各市场之间的关联比较密切,因此,该模型也可以用来模拟预测其他稀土氧化物的价格。该模型具有一定的实践运用价值,稀土行业管理部门可以运用该模型定期编制稀土产品价格预测报告,以便稀土产品生产经营者、消费者和各级政府随时掌握稀土产品市场价格情况和变动趋势,及时进行决策。
Due to the many factors affecting the price of rare earth products, the factors are intricately interrelated, and the monthly price data of rare earth products have a high degree of non-stationary, nonlinear and noise characteristics, which increases the difficulty of predicting the price of rare earth products. Therefore, using the time series forecasting method, ARIMA (1,1,2) model of non-stationary time series is established based on the monthly prices of rare earth Nd_2O_3 and Dy_2O_3 to describe and predict the dynamic changes of the prices of rare earth products. Month ~ December 2015, and the predictive accuracy of prediction of Nd_2O_3, Dy_2O_3 was calculated. The results show that the model has high fitting accuracy and is suitable for the prediction of Nd_2O_3, Dy_2O_3 products in the short to medium term price. Since the volatility cycles of the price of China’s major rare earth products have some similarities, indicating that the correlation between the markets is relatively close, the model can also be used to simulate and forecast the prices of other rare earth oxides. The model has some practical value, rare earth industry management department can use the model regular preparation of rare earth products price forecast report, so that rare earth products producers, consumers and governments at all levels keep abreast of rare earth products market prices and trends, timely Make a decision.