基于移动窗PLS的选矿产品质量指标建模  

A Moving Window PLS-based Modeling Approach for Product Quality Indicator in Mineral Processing

作　　者：余刚[1,2] 周俊武[1,2] 徐宁[1,2] 

机构地区：[1]北京矿冶研究总院,北京市100070 [2]矿冶过程自动控制技术北京市重点实验室,北京市100070

出　　处：《有色冶金设计与研究》2014年第4期4-7,共4页Nonferrous Metals Engineering & Research

基　　金：国家高技术研究发展计划(863计划)(2011AA060203)

摘　　要：通过分析选矿过程操作变量对精矿产品质量的影响,建立了一种基于移动窗PLS的精矿产品质量软测量模型,利用关键实时操作变量,对精矿质量指标进行预估,以某黄金选矿厂的精矿品位指标预测为例验证了该方法的有效性。The modeling of product quality is significant in achieving quality stability control and improving economic benefits for mineral processing plant. The influence of operating variables on concentrate quality are analyzed in the paper. A concentrate product quality soft sensor is proposed based on moving window PLS which applying the key real-time operating variables to predict the concentrate quality indicators. The experiment results of concentrate grade prediction for a gold mineral processing plant show the efficiency of this method.

关 键 词：质量指标建模 PLS 数据驱动 软测量 移动窗 

分 类 号：TD928.9[矿业工程—选矿]