摘要
文章分析了孔隙充水矿井的充水水源和通道,利用非线性的BP人工神经网络建立了徐州韩桥煤矿涌水量短期预测模型,选取每天的降水量作为影响因子,用已有的涌水量资料训练得到权值和阈值来表示充水通道,并对-200m水平、-270m水平、-330m水平和全矿井涌水量进行了预测。结果显示,涌水量的预测值与实测值吻合得较好,说明该模型具有一定实用性。
In this paper, sources and channels of water bursting of mine with pore water yield were analyzed and basic theory of artificial neural network was used. The short-time prediction model of mine inrush in the Hanqiao colliery was also established. Daily precipitation within a period of time was chosen as an influence factor. Weight and threshold, which were obtained from training known data of precipitation, were expressed as channels of water inrush. The mine inrush water of - 200 m level, - 270 m level, - 330 m level and the whole mine was predicted. The results show that it is right and feasible to build the BP neural network model and predict mine inrush water.
出处
《水文地质工程地质》
CAS
CSCD
北大核心
2007年第5期55-58,共4页
Hydrogeology & Engineering Geology
基金
国家自然科学基金重点项目"水资源保护性煤炭开采基础理论与应用研究"(50634050)
国家重点基础研究发展计划"973"计划(2007CB209401)
关键词
BP人工神经网络
孔隙充水矿井
涌水量
预测模型
韩桥煤矿
BP artificial neural network
mine with pore water yield
mine inrush water
prediction model
Hanqiao colliery
