基于时间序列模型的铀矿价格分析与预测

铀矿价格是其投资风险中最重要的不确定性因素，其价格预测的准确与否直接关系到企业投资的成败。根据1989-2009年铀矿价格数据，利用MATLAB数值分析软件建立时间序列模型，通过ES模型和DMA模型的预测数据与实际数据的比较，论证分析几种铀矿价格预测模型的科学性和实用性。结果表明：①平滑系数越大，标准误差越小；②4种不同方法中，三次指数平滑预测法效果最好，其模型计算结果与实际值基本一致；③在仅考虑价格时间序列本身因素的条件下，铀矿价格呈现逐渐增长的长期发展趋势。

Uranium Ore Price Analysis and Prediction Based on Time Series Model

Uranium ore price is the most important factors of the investment risk,the price forecast accurate or not is directly related to the investment success or failure.According to the uranium ore price data from 1989 to 2009,the time series mode is established based on MATLAB numerical analysis software,the comparison analysis of the prediction results of ES model and DMA model and the actual data is conducted to discuss the scientific and practicability of the time series model of several uranium ore prices,the results show that:①the greater the smoothing coefficient,the standard error of prediction is small;②among the four methods,the third-index flatness model is the best one,the calculated results of the model is basically consistent with the actual data;③under the condition that only the time series properties are considered,the uranium ore price can be gradually increasing.