全尾砂絮凝沉降参数优化研究  

Optimization on the Flocculation and Sedimentation Parameters of Total Tailings

作　　者：杨宁[1,2,3] 尹贤刚[1,2,3] 肖木恩[1,2,3] 赖伟[1,2,3] YANG Ning YIN Xiangang XIAO Muen LAI Wei(Changsha Institute of Mine Research Co., Ltd, Changsha, Hunan 410012, China State Key Laboratory of Safety Technology of Metal Mine, Changsha, Hunan 410012, China National Engineering Research Center for Metal Mining, Changsha, Hunan 410012, China)

机构地区：[1]长沙矿山研究院有限责任公司,湖南长沙410012 [2]金属矿山安全技术国家重点实验室,湖南长沙410012 [3]国家金属采矿工程技术研究中心,湖南长沙410012

出　　处：《矿业研究与开发》2017年第3期24-28,共5页Mining Research and Development

摘　　要：为使全尾砂絮凝沉降能够达到最佳的效果,运用BP神经网络对其沉降参数进行优化研究。该网络的输入因子包括絮凝剂的添加量q和全尾砂的浓度c,输出因子为尾砂的絮凝沉降速度v;以正交试验为手段,通过创建足够的网络学习样本,最终遴选出最适合的网络模型。将正交试验扩展,增进输入因子水平,把优选出的样本再次搭配,从而找寻到絮凝沉降参数的最优值。以某铅锌矿全尾砂的絮凝沉降为实例,优选出的q值为10g/t,c为15%,v为0.882m/h,能够满足现场要求。In order to achieve the best flocculation and sedimentation effect of total tailings, the BP neural network was used to optimize the parameters. The input factors of the network included the unit consumption of flocculant q and the concentration of total tailings c, and the output factor was only the flocculation and sedimentation velocity of tailings v. By the orthogonal experiments and through making enough network-learning samples, the most suitable network model was finally setected. Furthermore, by expanding the orthogonal experiments, the levels of input factors were improved, and the optimized samples were matched again to find out the best flocculation and sedimentation parameters. Taking the flocculation and sedimentation of total tailings in a lead-zinc mine as an example, the optimal parameters with q value of 10 g/t, c value of 15% and v value of 0.882 m/h could meet the field requirement. The application results showed that the research achieved significant effects, which provided a new idea for optimizing the flocculation and sedimentation parameters of total railings.

关 键 词：全尾砂 絮凝沉降 BP神经网络 

分 类 号：TD853.34[矿业工程—金属矿开采]