论文部分内容阅读
本文用经验的识别方法处理一种时间序列,这种时间序列是通过装在铅锌分离浮选回路上的传感器获得的。用ARMA模型(自回归滑动平均模型)拟合给矿锌品位,并用于预测给矿锌品位的动态变化。将尾矿和精矿的锌品位、及锌回收率作为三个输出变量,给矿品位和黄药添加速率作为二个输入变量,用三个传递函数描述该选别过程,然后用离线和在线的方法估计这些传递函数。
In this paper, an empirical method is used to process a time series obtained by a sensor mounted on a lead-zinc separation flotation circuit. The ARMA model (autoregressive moving average model) is used to fit the zinc ore grade and to predict the dynamic change of zinc ore grade. Tailings and concentrates of zinc grade, and zinc recovery rate as the three output variables, ore grade and xanthate addition rate as two input variables, using three transfer functions to describe the sorting process, and then offline and online The method to estimate these transfer functions.