基于数据优化的BP神经网络煤层底板破坏深度预测Coal Floor Failure Depth Prediction Based on Data Optimized BP Neural Network
薛喜成,田凡凡
摘要(Abstract)：
在煤矿安全生产工作中,准确预测工作面底板破坏深度对于预防工作面底板突水事故具有重要意义。本文基于数据优化的观点作了以下几方面工作:在分析煤层底板破坏深度发育因素的基础上,进行单因素方差分析,剔除了低相关度因素,采用熵权法得到了主控因素的权重占比;采用主成分分析法对剩余主控因素进行去冗余度处理;最后将主成分变量作为BP神经网络的输入端,建立底板破坏深度预测模型。结果表明:该模型预测结果的均方根误差(RMSE)仅为4.199,研究成果可为底板水害防治以及类似的方法提供一定的指导思路。
关键词(KeyWords)： 煤层底板破坏深度;方差分析;熵权法;BP神经网络
基金项目(Foundation):
作者(Author): 薛喜成,田凡凡
参考文献(References)：
- [1]李兴高，高延法.采场底板岩层破坏与损伤分析[J].岩石力学与工程学报，2003(1):35-39.Li X G,Gao Y Y.Damage analysis of floor strata[J].Chinese Journal of Rock Mechanics and Engineering,2003(1):35-39.
- [2]关英斌，李海梅，路军臣.显德汪煤矿9号煤层底板破坏规律的研究[J].煤炭学报，2003(2):121-125.Guan Y B,Li H M,Lu J C.Research of No.9 coal seam floor's fracture regularityin Xiandewang Coal Mine[J].Journal of China Coal Society,2003(2):121-125.
- [3]国家煤炭工业局.建筑物、水体、铁路及主要井巷煤柱留设与压煤开采规程[M].北京：煤炭工业出版社，2008.
- [4]施龙青，徐东晶，邱梅，等.采场底板破坏深度计算公式的改进[J].煤炭学报，2013,38(S2):299-303.Shi L Q,Xu D J,Qiu M,et al.Improved on the formula about the depth of damaged floor in working area[J].Journal of China Coal Society[J].Journal of China Coal Society,2013,38(S2):299-303.
- [5]程久龙，于师建，宋扬，等.煤层底板破坏深度的声波CT探测试验研究[J].煤炭学报，1999(6):576-580.Cheng J L,Yu S J,Song Y,et al.Detection of the failure depth of coal seam floor by acoustic wave computer tomography[J].Journal of China Coal Society,1999(6):576-580.
- [6]吕春峰，王芝银，李云鹏.含裂隙煤层底板突水规律的数值模拟与工程应用[J].岩土力学，2003(S2):112-116.Lv C F,Wang Z Y,Li Y P.Numerical simulation of water inrush law in coal seam floor with cracks and its application to engineering practice[J] Rock and Soil Mechanics,2003(S2):112-116.
- [7]于小鸽，韩进，施龙青，等.基于BP神经网络的底板破坏深度预测[J].煤炭学报，2009,34(6):731-736.Yu X G,Han J,Shi L Q,et al.Forecast of destroyed floor depth based on BP neural networks[J] Journal of China Coal Society,2009,34(6):731-736.
- [8]张文泉，赵凯，张贵彬，等.基于灰色关联度分析理论的底板破坏深度预测[J].煤炭学报，2015,40(S1):53-59.Zhang W Q,Zhao K,Zhang G B,et al.Prediction of floor failure depth based on grey correlation analysis theory[J].Journal of China Coal Society,2015,40(S1):53-59.
- [9]路畅，尹立明，李杨杨，等.基于灰色系统理论的煤层底板破坏深度预测模型[J].矿业研究与开发，2015,35(6):63-67.Lu C,Li L M,Li Y Y,et al.Prediction Model of Destroyed Floor Depth Based on Grey System Theory[J].Mining Research and Development,2015,35(6):63-67.
- [10]施龙青，张荣遨，韩进，等.基于GWO改进的PCA-BP神经网络煤层底板破坏深度预测模型[J].矿业研究与开发，2020,40(2):88-93.Shi L Q,Zhang R A,Han J,et al.The depth prediction model of PCA-BP neural network based on GWO[J].Mining Research and Development,2020,40(2):88-93.
- [11]张思建.综采工作面底板破坏深度研究[J].煤矿现代化，2018(5):166-167.Zhang S J.Study on the depth of damaged floor of fully mechanized working face[J].Coal Mine Modernization,2018(5):166-167.
- [12]李江华，许延春，谢小锋，等.采高对煤层底板破坏深度的影响[J].煤炭学报，2015,40(S2):303-310.Li J H,Xu Y C,Xie X F,et al.Influence of mining height on coal seam floor failure depth[J].Journal of China Coal Society,2015,40(S2):303-310.
- [13]刘伟韬，穆殿瑞，谢祥祥，等.倾斜煤层底板采动应力分布规律及破坏特征[J].采矿与安全工程学报，2018,35(4):756-764.Liu W T,Mu D R,Xie X X,et al.Rules of stress distributions and failure characteristics in inclined coal seam floor due to mining[J].Journal of Mining & Safety Engineering,2018,35(4):756-764.
- [14]张风达，申宝宏.深部煤层底板破坏特征分析[J].采矿与安全工程学报，2019,36(1):44-50.Zhang F D,Sheng B H.Failure characteristics analysis of deep coal seam floor[J].Journal of Mining & Safety Engineering,2019,36(1):44-50.
- [15]郝哲，刘斌.基于差分法及神经网络的硐室围岩力学参数反分析[J].岩土力学，2003(S2):77-80.Hao Z,Liu B.Back analysis of mechanics parameters of rocks surrounding openings on the basis of calculus of difference and neural network[J].Rock and Soil Mechanics,2003(S2):77-80.