Flotation Recovery Prediction of Zijinshan Copper Ore Based on I−GWO−BP Neural Network
-
摘要:
为克服传统测量浮选回收率方式存在的低效率、滞后性等问题，结合紫金山硫化铜矿浮选厂生产情况，采用基于MI（Mutual Information）互信息法对选厂原矿品位、丁铵黑药用量等浮选条件因子进行特征选择，在此基础上，建立了基于BP（Back Propagation）、GWO−BP（Grey Wolf Optimizer−Back Propagation）、I−GWO−BP（Improved−Grey Wolf Optimizer−Back Propagation）的三种浮选回收率预测模型，并选取紫金山硫化铜矿浮选车间生产数据进行神经网络训练与验证试验，分析了浮选回收率预测模型的准确性。结果表明：相较于基于BP、GWO−BP的浮选回收率预测模型而言，基于I−GWO−BP的浮选回收率预测模型具有更大的相关系数和更小的均方误差根，说明该模型泛化拟合能力更强，对浮选回收率的预测值在很大程度上逼近于真实值，预测精度更高。本研究结果可为实现浮选回收率高效、准确、自动的在线预测技术开发提供支持。
Abstract:The traditional method of measuring flotation recovery has some problems, such as low efficiency and hysteresis. Combined with the flotation plant production of Zijinshan sulfide copper ore, characteristics selection of flotation condition factors such as raw ore grade and dosage of ammonium dibutyl dithiophosphate was carried out based on MI (Mutual Information) method. On this basis, three prediction models of flotation recovery were established based on BP (Back Propagation) GWO−BP (Grey Wolf Optimizer−Back Propagation) and I−GWO−BP (Improved−Grey Wolf Optimizer−Back Propagation) . The flotation workshop production data of Zijinshan sulfide copper ore were selected for neural network training and verification test, and the accuracy of the flotation recovery prediction model was analyzed. The results showed that compared with BP and GGO−BP, the flotation recovery prediction model based on I−GWO−BP had a root mean squared error and a correlation coefficient and the predicted value of flotation recovery was the closest to the true value, and the generalization ability of the network was significantly stronger. The results of this study can support the development of efficient, accurate and automatic online prediction techniques for flotation recovery.
-
-
表 1 不同隐含层节点数对应的网络均方误差
Table 1. Network mean squared error corresponding to the number of nodes in different hidden layers
隐含层节点数 相应的均方误差 3 0.037566 4 0.035899 5 0.036765 6 0.031699 7 0.029845 8 0.03125 9 0.035295 10 0.029027 11 0.031674 12 0.029306 
下载: 导出CSV
表 2 BP神经网络的基本参数表
Table 2. Basic parameters of BP neural network
网络参数名 参数值 网络参数名 参数值 输入层节点 4 输出层激活函数 Purelin 函数 输出层节点 1 最大迭代次数 1000 隐含层节点 10 学习率 0.1 学习函数 LM算法 学习精度 0.0001 隐含层激活函数 Logsig函数
下载: 导出CSV
表 3 不同网络模型对应的验证实验评价结果
Table 3. Evaluation results of validation experiments corresponding to different network models
网络名称 R值 RMSE值 MAE值 MAPE值 BP 0.73 1.8415 1.0187 1.4274% GWO−BP 0.75 1.8219 1.2161 1.2243% IGWO−BP 0.94 0.2210 0.1856 0.2067%
下载: 导出CSV
-
[1] 刘青, 王彬, 袁玮, 等. 金矿浮选回收率预测模型[J]. 北京科技大学学报, 2014, 36(11): 1456−1461.
LIU Q, WANG B, YUAN W, et al. Prediction model of flotation recovery from gold mine[J]. Journal of University of Science and Technology Beijing, 2014, 36(11): 1456−1461.
[2] 刘利敏, 杨文旺, 刘之能, 等. 基于BP神经网络的浮选回收率预测模型[J]. 有色金属(选矿部分), 2013(S1): 206−208.
LIU L M, YANG W W, LIU Z N, et al. Flotation recovery prediction model based on BP neural network[J]. Nonferrous Metals (Mineral Processing Section), 2013(S1): 206−208.
[3] 王衡嵩, 魏志聪, 彭蓉, 等. 基于RSM和RA−BPNN的锌窑渣中铜浮选试验优化[J]. 人工晶体学报, 2019, 48(8): 1557−1564. doi: 10.3969/j.issn.1000-985X.2019.08.030
WANG H S, WEI Z C, PENG R, et al. Optimization of copper flotation test in zinc kiln slag based on RSM and RA−BPNN[J]. Journal of Synthetic Crystals, 2019, 48(8): 1557−1564. doi: 10.3969/j.issn.1000-985X.2019.08.030
[4] 张勇, 朱晶. 基于混沌蚁群神经网络的浮选过程经济技术指标预测[J]. 上海交通大学学报, 2016, 50(6): 975−979.
ZHANG Y, ZHU J. Prediction of economic and technical indicators of flotation process based on chaotic ant colony neural network[J]. Journal of Shanghai Jiaotong University, 2016, 50(6): 975−979.
[5] 冉宇, 李梅, 高凯, 等. 稀土矿浮选泡沫大小与回收率的相关性[J]. 化工矿物与加工, 2019, 48(11): 24−28. doi: 10.16283/j.cnki.hgkwyjg.2019.11.007
RAN Y, LI M, GAO K, et al. Correlation between flotation foam size and recovery rate of rare earth ore[J]. Chemical Minerals and Processing, 2019, 48(11): 24−28. doi: 10.16283/j.cnki.hgkwyjg.2019.11.007
[6] MIRJALILI S M, LEWIS A. Grey wolf optimizer[J]. Advances in Engineering Software, 2014, 69: 46−61. doi: 10.1016/j.advengsoft.2013.12.007
[7] 吴丁杰, 周庆兴, 温立书. 基于Logistic混沌映射的改进麻雀算法[J]. 高师理科学刊, 2021, 41(6): 10−15. doi: 10.3969/j.issn.1007-9831.2021.06.003
WU D J, ZHOU Q X, WEN L S. Improved sparrow algorithm based on logistic chaos mapping[J]. Journal of Science of Teachers' College and University, 2021, 41(6): 10−15. doi: 10.3969/j.issn.1007-9831.2021.06.003
[8] 刘公致, 吴琼, 王光义, 等. 改进型Logistic混沌映射及其在图像加密与隐藏中的应用[J]. 电子与信息学报, 2022, 44(10): 3602−3609.
LIU G Z, WU Q, WANG G Y, et al. Improved Logistic chaos mapping and its application in image encryption and concealment[J]. Journal of Electronics & Information Technology, 2022, 44(10): 3602−3609.
[9] 马晓宁, 李笑含. 基于Tent混沌映射的可复制的鲸鱼算法[J]. 计算机仿真, 2022, 39(8): 363−368. doi: 10.3969/j.issn.1006-9348.2022.08.069
MA X N, LI X H. Reproducible whale algorithm based on Tent chaos mapping[J]. Computer Simulation, 2022, 39(8): 363−368. doi: 10.3969/j.issn.1006-9348.2022.08.069
[10] 刘园园, 贺兴时. 基于Tent混沌映射的改进的萤火虫算法[J]. 纺织高校基础科学学报, 2018, 31(4): 511−518.
LIU Y Y, HE X S. Improved firefly algorithm based on Tent chaos mapping[J]. Journal of Basic Science of Textile Universities, 2018, 31(4): 511−518.
[11] 马良玉, 於世磊, 王佳聪, 等. 基于XGBoost建模及改进灰狼优化算法的再热汽温预测优化控制[J]. 热能动力工程, 2023, 38(1): 104−111.
MA L Y, YU S L, WANG J C, et al. Optimization control of reheat steam temperature prediction based on XGBoost modeling and improved gray wolf optimization algorithm[J]. Journal of Engineering for Thermal Energy and Power, 2023, 38(1): 104−111.
-
图(8)
表(3)
计量
- 文章访问数: 570
- PDF下载数: 94
- 施引文献: 0