首页|期刊导航|中国钼业|基于支持向量机模型的某铜矿回收率预测研究

基于支持向量机模型的某铜矿回收率预测研究OA

Prediction of recovery rate in a copper mine based on a support vector machine model

中文摘要英文摘要

为探究某铜矿粗选工艺中各因素对铜回收率的影响规律并建立精确预测模型,开展了黄铜矿浮选的单因素与正交试验,并采用支持向量机(SVM)对铜回收率进行预测.单因素试验表明:在Na2S用量110 g/t、丁基黄药55 g/t、酯205用量28 g/t、2#油用量45.5 g/t的条件下,粗精矿中铜品位和回收率分别达到7.15%和95.23%.通过正交试验与极差分析筛选出关键影响因素,并构建SVM预测模型,结果显示:模型训练拟合优度R2为0.9887,拟合效果良好;16组验证试验中预测回收率与实际回收率的平均绝对误差为0.117 3%,预测精度较好.研究表明:SVM模型能够有效预测铜回收率,对某黄铜矿粗选浮选过程的优化与回收率预测具有指导意义.

To investigate the influence of various factors in the copper roughing process on copper recovery and establish an accurate prediction model,single-factor and orthogonal experiments were conducted on chalcopyrite flotation.A support vector machine(SVM)model was applied to predict the copper recovery.Single-factor tests showed that under the conditions of Na2S dosage of 110 g/t,butyl xanthate of 55 g/t,ester-205 of 28 g/t,and 2# oil of 45.5 g/t,the copper grade and recovery in the rough concentrate reached 7.15%and 95.23%,respectively.Key influencing factors were identified through orthogonal experiments and range analysis,and an SVM pre-diction model was constructed.The results indicated that the trained model achieved a coefficient of determination(R2)of 0.988 7,good fitting performance.The mean absolute error between the predicted and actual recovery values in 16 validation tests was 0.117 3%,indicating that the prediction accuracy meets the required standards.This study confirms that the SVM model can effectively predict copper recovery,providing valuable guidance for optimizing the flotation process and predicting recovery in chalcopyrite roughing opera-tions.

彭祥玉;赵椒嵘;张钰;吴忠意;王宇斌

西安建筑科技大学资源工程学院,陕西西安 710055西安建筑科技大学资源工程学院,陕西西安 710055西安建筑科技大学资源工程学院,陕西西安 710055西安建筑科技大学资源工程学院,陕西西安 710055西安建筑科技大学资源工程学院,陕西西安 710055

矿业与冶金

黄铜矿SVM模型铜回收率浮选

chalcopyriteSVM modelcopper recoveryflotation

《中国钼业》 2026 (4)

60-65,6

陕西省科技厅青年基金(编号:2024JC-YBQN-0389)陕西省科学技术协会青年人才托举计划项目(编号:20250403)

10.13384/j.cnki.cmi.1006-2602.2026.04.008

评论