针对金属矿山井下粉尘来源分析中指标参数的不确定性，为提高算法收敛速度和准确度，提出了基于SOFM神经网络的方法；以铅锌矿为例，据此建立了铅锌矿井下粉尘来源状态的神经网络模型。采用收集到的铅锌矿井下粉尘样本进行实例分析，运用Matlab软件对该模型进行了训练和检验，并与其他竞争型神经网络进行了对比分析。其分析结果表明：与竞争型神经网络相比，SOFM神经网络的性能更优、泛化能力更强、预测精度更高，是金属矿山井下粉尘来源可靠性分析的一种有效方法。
In view of the uncertainty of indicator parameters in the analysis of dust source in underground metal mines,and in order to improve the convergence speed and accuracy,this paper puts forward a method based on SOFM neural network.With lead-zinc mine as the research subject,a neural network model of underground dust source is established based on it.The collected sample of mine dusts is empirically analyzed.Matlab is used to train and test the model.And comparison is made with other competitive neural network.The analysis shows that compared with other competitive networks,SOFM neural network performs better,applies more widely,predicts more precisely and is an ef-ficient way for reliable analysis of underground dust source in metal mines.
参考文献
| [1] | 王晓雪;王林山.基于SOFM神经网络的学生综合评价[J].河北师范大学学报(自然科学版),2011(3):239-243. |
| [2] | 张蓉 .铅锌矿矿井污风排放引发局部重金属污染的机制研究[D].中南大学,2014. |
| [3] | 杜之正.基于BP神经网络电除尘效率模型研究[J].东北电力技术,2014(09):8-10,13. |
| [4] | 马程;李双成;刘金龙;高阳;王阳.基于SOFM网络的京津冀地区生态系统服务分区[J].地理科学进展,2013(9):1383-1393. |
| [5] | 黄姣;高阳;赵志强;李双成.基于GIS与SOFM网络的中国综合自然区划[J].地理研究,2011(9):1648-1659. |
| [6] | 倪步喜;章丽芙;姚敏.基于SOFM网络的聚类分析[J].计算机工程与设计,2006(5):855-856,878. |
| [7] | 张设计;刘勇;周润金;郑彦奎.掘进工作面粉尘分布规律及控降尘工艺技术试验[J].矿业安全与环保,2010(2):30-33. |
- 下载量()
- 访问量()
- 您的评分：
-
10%
-
20%
-
30%
-
40%
-
50%