![c9603be79d33e0f62a98120594e86dfd.png](https://i-blog.csdnimg.cn/blog_migrate/5cf5183d78f73a2a0a708f126ba91e55.png)
点击蓝字关注地下空间
![c9603be79d33e0f62a98120594e86dfd.png](https://i-blog.csdnimg.cn/blog_migrate/5cf5183d78f73a2a0a708f126ba91e55.png)
Stability analysis of underground mine hard rock pillars via combination of finite difference methods, neural networks, and Monte Carlo simulation techniques
基于有限差分法、神经网络和蒙特卡罗模拟技术的井下硬岩矿柱稳定性分析
Underground Space, Available online 19 June 2020
https://doi.org/10.1016/j.undsp.2020.05.005
![0c3f66061dd3b30d9ad6d2f40ce5c723.png](https://i-blog.csdnimg.cn/blog_migrate/2aeb8e2056071875c14439c9c085597a.png)
求解矿柱应力的反向传递神经网络结构
Proposed BPNN structure for solving pillar stress
Authors
作者
Chuanqi Li(a), Jian Zhou(b✉*), Danial Jahed Armaghani(c), Xibing Li(d)
a. School of Resources and Safety Engineering, Central South University, Changsha 410083, China
b. School of Resources and Safety Engineering, Central South University, Changsha 410083, China
c. Department of Civil Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia
d. School of Resources and Safety Engineering, Central South University, Changsha 410083, China
✉ csujzhou@hotmail.com
Abstract
摘要
Pillar stability is always evaluated using the safety factor (SF), which is defined as the ratio of pillar strength to pillar stress. However, most researchers have estimated pillar stress using the pillar shape (w/h), uniaxial compressive strength (UCS) of the intact rock mass, and pillar depth (H). In this study, the geological strength index (GSI) of hard rock pillars was considered as a new variable for predictive purposes. This index was developed by combining numerical simulation software (i.e., FLAC 3D) and a back-propagation neural network (BPNN). A hard rock pillar stability analysis, based on three methods: deterministic, sensitivity analysis, and Monte Carlo simulation (MCS), was performed. A new formula was proposed to estimate the SF values based on the predicted stress, considering the GSI variable in the deterministic method. The sensitivity analysis indicated that the variables impacting the SF from high to low are UCS, GSI, w/h, and H. In this study, pillar stability was analyzed mainly using the GSI and MCS techniques. The MCS results revealed that the GSI is also a major factor in pillar stability and has a greater effect on weak pillars than on strong ones. Furthermore, a pillar is more likely to be unstable when both the GSI and the UCS (GSI-UCS) are decreased. This study provides several references and procedures for improving the design of stable pillars considering the GSI as an important factor.
矿柱稳定性通常采用安全系数(SF)进行评估,安全系数定义为矿柱强度与矿柱应力之比。然而,大多数研究者都是利用柱体形状(w/h)、完整岩体的单轴抗压强度(UCS)和矿柱深度(h)来估算矿柱应力。在本研究中,硬岩矿柱的地质强度指数(GSI)被视为一个新的预测变量。该指数是结合数值模拟软件(即flac3d)和bp神经网络(BPNN)得到的。基于确定性分析、敏感性分析、蒙特卡罗模拟三种方法进行了岩石稳定性分析。考虑确定性方法中的GSI变量,提出了一种基于预测应力的安全系数值估计公式。敏感性分析表明,影响矿柱稳定性的因素由高到低依次为UCS、GSI、w/h和h。MCS结果表明,GSI也是矿柱稳定性的主要因素,其对弱矿柱的影响大于对强矿柱的影响。此外,当GSI和UCS(GSI-UCS)都降低时,支柱更可能不稳定。本研究为考虑GSI的更好的稳定矿柱设计提供参考资料及流程指导。
Keywords
关键词
Hard rock pillar; Numerical simulation; Neural networks; Safety factor; Geological strength index; Monte Carlo simulation
硬岩矿柱; 数值模拟; 神经网络; 安全系数; 地质强度指标; 蒙特卡罗模拟
![c9603be79d33e0f62a98120594e86dfd.png](https://i-blog.csdnimg.cn/blog_migrate/5cf5183d78f73a2a0a708f126ba91e55.png)
![a2007055e20dcca58aee53225f676db1.png](https://i-blog.csdnimg.cn/blog_migrate/6d0a84406a4893bfbad5506201836802.jpeg)
Underground Space
微信号:UNDSP_Journal
点击阅读原文查看文章详情
▇ 扫码关注我们
![c9603be79d33e0f62a98120594e86dfd.png](https://i-blog.csdnimg.cn/blog_migrate/5cf5183d78f73a2a0a708f126ba91e55.png)