全尾砂充填体正交-BP神经网络强度预测

发布时间：2018-05-07 15:05

本文选题：全尾砂 + 减水剂　；参考：《金属矿山》2016年01期

【摘要】：某矿山采用全尾砂胶结充填,需加入减水剂以保证充填料浆流动性能,其充填体强度各影响因素之间存在着更为复杂的化学物理作用。为了解各因素对充填体强度的影响规律,准确预测其强度,建立了基于正交试验的BP神经网络全尾砂胶结充填体强度预测模型。预测过程中,为提高样本可信度,以料浆浓度、灰砂比、减水剂比例为影响因素,设计了3因素、4水平正交试验方案作为研究数据基础。采用灰色关联度理论分析了各因素对全尾砂充填体强度的影响规律,结果表明:对充填体早期强度(7 d)影响最大的是料浆质量浓度,其次为灰砂比、减水剂比例,对充填体后期强度(28 d)影响最大的是灰砂比,其次为料浆浓度、减水剂。以7 d、28 d强度为输出因子,运用BP神经网络对充填体强度进行预测,预测结果与试验结果最大误差为9.98%,平均误差为2.71%,精度较高,预测可靠性强,具有较好的工程应用价值。
[Abstract]:In order to ensure the fluidity of the filling slurry, the cemented filling of the whole tailings is adopted in a mine, and the influence factors of the backfill strength have more complex chemical and physical effects. The addition of water reducing agent is necessary to ensure the fluidity of the filling slurry. In order to understand the influence of various factors on the strength of the filling body and predict its strength accurately, a BP neural network model for predicting the strength of the cemented filling body was established based on the orthogonal test. In order to improve the reliability of the samples, three factors, namely, slurry concentration, ratio of ash to sand and ratio of water reducer, were used as the data basis of the study. The influence of various factors on the strength of the whole tailings filling body is analyzed by using the grey correlation degree theory. The results show that the mass concentration of slurry is the most important factor, followed by the ratio of lime to sand and the ratio of water reducer to water reducing agent, which has the greatest influence on the early strength of the filling body for 7 days. The ratio of ash to sand is the most important influence on the later strength of the filling body, followed by slurry concentration and water reducing agent. The maximum error between the prediction result and the test result is 9.98, the average error is 2.71, the precision is high, the prediction reliability is strong, and it has good engineering application value, taking the intensity of 7 days and 28 days as the output factor, using BP neural network to predict the strength of the filling body, the maximum error between the prediction result and the test result is 9.98 and the average error is 2.71%.
【作者单位】：新桥硫铁矿;中南大学资源与安全工程学院;
【基金】：“十二五”国家科技支撑计划项目(编号:2012BAC09B02)
【分类号】：TD853.34

【相似文献】