原地爆破浸出率的时间序列预测问题初探

采用灰色系统理论对原地爆破浸出率的时间序列预测问题进行了研究。根据原地爆破浸出工艺的浸出动力学特点，结合现场实例，建立了某铀矿山原地爆破浸出率时间序列预测的短期等维新息GM(1，1)模型和长期等维灰数递补GM(1，1)模型。分别采用这两个模型进行了短期和长期预测，其最大相对误差分别为4．4％和10％，预测结果表明：预测方法合理，预测精度可靠。

Primary Discussion on the Time Series Prediction of the Leaching Rate of In-situ Blasting and Leaching Ore

In this paper, the theory of gray system is adopted to study the time series prediction of the leaching rate of in-situ blasting and leaching ore. According to the dynamic characteristics of in-situ leaching process and combined with the practical engineering of a certain uranium mine, the GM(1,1) models of equal dimension new signal and equal dimension successive gray data compensation are established respectively for the short-term and long-term time prediction of the leaching rate of in-situ blasting and leaching ore in the mine, and the short-term and long-term time prediction of the leaching rate of the example mine are carried out, the maximum relative errors of predicted values of this two models are 4.4% and 10%. The prediction results show that the adopted approach is reasonable and prediction accuracy is reliable.