| 改进的AHP结合BP神经预测煤与瓦斯突出 |
| |
| 引用本文： | 周西华,郭坤,白刚,宋东平.改进的AHP结合BP神经预测煤与瓦斯突出[J].物探化探计算技术,2019(1):121-127. |
| |
| 作者姓名： | 周西华 郭坤 白刚 宋东平 |
| |
| 作者单位： | 辽宁工程技术大学安全科学与工程学院;矿山热动力灾害与防治教育部重点实验室;辽宁工程技术大学煤炭资源安全开采与洁净利用工程研究中心 |
| |
| 基金项目： | 辽宁省教育厅基金(551610001276;L2012122) |
| |
| 摘 要： | 为对煤与瓦斯突出进行准确预测,结合重要性指标筛选以及改进的层次分析法,对权重值进行研究,并进行排序。通过筛选重要性指标,选定其中具有突出影响的指标。结果表明,对此造成影响的因素权重分别为地质构造、地应力、瓦斯压力、放散初速度。未经筛选的BP神经网络的实际值与预测值的平均相对误差为3.44%,而通过筛选后选定的指标,实际值与预测值的平均相对误差为1.82%,训练速度、精度优于未经筛选指标的预测。可为防突措施的奠定理论基础,对煤矿安全生产工作具有借鉴作用。
|
| 关 键 词： | 层次分析法 重要性指标 筛选 BP神经 |
Improved AHP combined with BP neural prediction of coal and gas outburst |
| |
| ZHOU Xihua,GUO Kun,BAI Gang,SONG Dongping.Improved AHP combined with BP neural prediction of coal and gas outburst[J].Computing Techniques For Geophysical and Geochemical Exploration,2019(1):121-127. |
| |
| Authors: | ZHOU Xihua GUO Kun BAI Gang SONG Dongping |
| |
| Institution: | (College of Safety Science and Engineering,Liaoning Technical University,Fuxin123000,China;Key Laboratory of Mine Thermodynamic Disasters and Control of Ministry of Education,Fuxin123000,China;Funded by Open Projects of Resarch Center of Coal Resources Safe Mining and Clean Utilization,LNTU,Fuxin123000,China) |
| |
| Abstract: | In order to predict the coal and gas outburst accurately,the weight value is studied and sorted by the importance index screening and the combination of improved analytic hierarchy process.Through screening the importance of indicators,it has a prominent impact on the selection of indicators.The results show that the factors influencing the factors are the geological structure,the geostress,the gas pressure and the initial velocity.The average relative error of the uncovered BP neural network is 3.44%,and the average relative error of the actual value and predicted value is 1.97%,the training speed is high and the prediction precision is high.This can lay a theoretical foundation for the formulation of outburst prevention measures and provide an evaluation method which is consistent with the actual situation. |
| |
| Keywords: | analytic hierarchy process importance index screening BP neural |
| 本文献已被 CNKI 维普 等数据库收录！ |
|
