摘要
针对某尾矿库在实际运行管理中需要快速判断库体稳定的实际需要,在仔细分析尾矿坝失稳原因的基础上结合赵家沟的工程实际情况确定了浸润线埋深、退坡距离、干滩长度、粘聚力、摩擦力等5个主要的影响因素,通过正交试验设计方法设计了不同的试验组合,利用GEO-Slope边坡稳定计算分析软件计算得到各因素组合下安全系数作为样本数据。根据改进的神经网络原理建立了影响因素与安全系数间的非线性映射网络模型,通过对样本数据的训练得到预测模型,并将该模型用于安全系数的预测,通过实际值与预测值的对比表明所建立的预测模型具有较高的精确度,可以用于尾矿坝坝坡稳定预测。利用本文所建立的尾矿坝坝坡稳定预测模型,通过给定影响因素值,可快速得到坝坡稳定安全系数,计算方便,可用于指导尾矿坝填筑单位在实际生产运行过程中对堆积坝填筑的控制,保障尾矿坝安全。
In view of the actual needs of a tailings dam,it was necessary to judge the stability of the tailings dam quickly during the actual operation and management.Based on the thorough analysis of the instability cause of the tailings dam,and combined the actual situation of Zhaojiagou,five main influencing factors were determined,including the depth of infiltration line,the distance of slope,the length of dry beach,cohesion and friction.On this basis,different test combinations were designed by orthogonal experimental design method,and the safety factor of each factor combination was calculated as sample data by GEOslope Slope.According to the principle of the improved BP neural network,a nonlinear mapping network model between influencing factors and safety factors was established.The prediction model was obtained through the training of sample data and applied to the prediction of safety factor.By comparing the actual values and the predicted values,it showed that the model had high accuracy and could be used to predict the stability of the tailings dam.Using the established model in this paper,the safety factor of the dam slope could be quickly obtained by given the influencing factor value,which was convenient and could be used to guide the filling process of the tailings dam during the actual production operation and ensure the safety of tailings dam.
作者
张圣
李同春
周桂云
晁阳
ZHANG Sheng;LI Tongchun;ZHOU Guiyun;CHAO Yang(College of Water Conservancy and Hydropower Engineering,Hohai University,Nanjing,Jiangsu 210098,China;College of Agricultural Science And Engineering,Hohai University,Nanjing,Jiangsu 210098,China;Jinling Institute of Technology,Nanjing,Jiangsu 21169,China)
出处
《矿业研究与开发》
CAS
北大核心
2019年第1期72-76,共5页
Mining Research and Development
基金
国家重点研发计划项目(2016YFCC00601)
关键词
尾矿坝
正交试验设计
改进神经网络
安全系数
Tailings dam
Orthogonal experimental design
Improved neural network
Safety factor
