基于MATLAB的锡石多金属硫化矿磨矿产品粒度分布预测研究  

Prediction of Grinding Product Size Distribution of Cassiterite Polymetallic Sulfide Ore Based on MATLAB

作　　者：朱朋岩 杨金林 马少健 蒋林伶 帅智超 ZHU Pengyan;YANG Jinlin;MA Shaojian;JIANG Linling;SHUAI Zhichao(College of Resources,Environment and Materials,Guangxi University,Nanning 530004,China;Guangxi Key laboratory of Processing for Nonferrous Metallic and Featured Materials,Nanning 530004,China)

机构地区：[1]广西大学资源环境与材料学院,南宁530004 [2]广西有色金属及特色材料加工重点实验室,南宁530004

出　　处：《有色金属（选矿部分）》2022年第2期66-72,共7页Nonferrous Metals（Mineral Processing Section）

基　　金：国家自然科学基金资助项目(51874105);广西自然科学基金资助项目(2018GXNSFAA281204)。

摘　　要：对磨矿产品粒度分布进行精准预测是实现磨矿智能化调控和优化的有效途径之一。基于锡石多金属硫化矿磨矿试验数据和MATLAB编程技术,进行了建立粒子群算法BP神经网络磨矿产品粒度分布预测模型研究。结果表明,在不同直径球介质和不同磨矿时间条件下,除少数几个粒级的平均相对误差在1%~2.22%,其他均在0.6%以下,预测结果比较理想。这说明预测模型的预测值与试验值吻合度高,模型可靠性高、适用性好。研究结果为锡石多金属硫化矿磨矿产品粒度分布预测提供一种新方法,也为锡石多金属硫化矿磨矿智能化调控和优化控制提供理论基础。Accurate prediction of particle size distribution of grinding products was one of the effective ways to realize intelligent regulation and optimization of grinding.Based on the grinding test data of cassiterite polymetallic sulfide ore and MATLAB programming technology,the particle swarm algorithm BP neural network prediction model of grinding product size distribution had been established.The results showed that the average relative error of a few sizes was 1%-2.22%,and the others was less than 0.6%under the conditions of different diameter ball medium and different grinding time.This indicated that the predicted value of the prediction model was in good agreement with the experimental value,and the model had high reliability and good applicability.The research results provided a new method for predicting the particle size distribution of cassiterite polymetallic sulfide grinding products,and also provided a theoretical basis for intelligent regulation and optimal control of cassiterite polymetallic sulfide grinding.

关 键 词：锡石多金属硫化矿 磨矿 MATLAB 粒度分布 预测 

分 类 号：TD921.4[矿业工程—选矿]