阴山学刊(自然科学版)
陰山學刊(自然科學版)
음산학간(자연과학판)
Yinshan Academic Journal
2015年
4期
9-12,19
,共5页
灰色神经网络%关联度%底板破坏深度%BP 算法
灰色神經網絡%關聯度%底闆破壞深度%BP 算法
회색신경망락%관련도%저판파배심도%BP 산법
Grey -neural network%Correlation degree%Failure depth of coal seam floor%BP algorithm
煤层底板破坏深度是煤矿底板突水研究中的重要参数之一,较为准确预测该参数可减少突水事故的发生。本文首先对煤层底板破坏深度的影响因素,按照关联度的大小排序进行了灰色关联定量分析,筛选出4个主控因子作为神经网络的输入元,同时应用优化 BP 算法,选用国内15个典型实测数据进行网络训练学习,建立灰色-神经网络预测模型,并以两淮矿区5个煤矿工作面的实例数据进行模型精准度检验,检验结果显示实测数据与预测结果有很好的一致性。最后将该模型应用袁店煤矿1011工作面,预测其煤层底板破坏深度值为14.23m。因此,该方法精度较高,适用于煤层底板破坏深度的预测。
煤層底闆破壞深度是煤礦底闆突水研究中的重要參數之一,較為準確預測該參數可減少突水事故的髮生。本文首先對煤層底闆破壞深度的影響因素,按照關聯度的大小排序進行瞭灰色關聯定量分析,篩選齣4箇主控因子作為神經網絡的輸入元,同時應用優化 BP 算法,選用國內15箇典型實測數據進行網絡訓練學習,建立灰色-神經網絡預測模型,併以兩淮礦區5箇煤礦工作麵的實例數據進行模型精準度檢驗,檢驗結果顯示實測數據與預測結果有很好的一緻性。最後將該模型應用袁店煤礦1011工作麵,預測其煤層底闆破壞深度值為14.23m。因此,該方法精度較高,適用于煤層底闆破壞深度的預測。
매층저판파배심도시매광저판돌수연구중적중요삼수지일,교위준학예측해삼수가감소돌수사고적발생。본문수선대매층저판파배심도적영향인소,안조관련도적대소배서진행료회색관련정량분석,사선출4개주공인자작위신경망락적수입원,동시응용우화 BP 산법,선용국내15개전형실측수거진행망락훈련학습,건립회색-신경망락예측모형,병이량회광구5개매광공작면적실례수거진행모형정준도검험,검험결과현시실측수거여예측결과유흔호적일치성。최후장해모형응용원점매광1011공작면,예측기매층저판파배심도치위14.23m。인차,해방법정도교고,괄용우매층저판파배심도적예측。
The failure depth of coal seam floor is one of the important parameters in the course of water burst-ing,so more accurately predict the parameters can reduce the water bursting accidents.This paper applied grey correlation analysis to quantitatively analyze the influence factors and got the correlation degree sort of each factor by sorting of correlation degree.Choosing four dominant factors as the input layers,grey -neural forecasting network has been built.Applying neural network algorithm,this model was trained by a total of 15 typical measured data collected from working face all over China,and 5 examples from coal mines of Huaibei&Huainan coalfield were used to test the practicability and the accuracy of model.The results indicate that high consistency between predic-ted value and measured values.Finally,applying the model in 1011 working face of Yuandian coal mine,the pre-dicted value is 14.23m.So this method can satisfy the precision requirement and can be used to predict failure depth of coal seam floor.