矿井通风安全数据智能分析研究
张超
（山西煤炭进出口集团左云东古城煤业有限公司，山西 左云 037100）
摘要： 以东古城煤业大量通风监测数据为基础，建立了通风数据算法模型，通过MGM(1，n)灰色预测模型和BP 神经
网络模型预测了瓦斯体积分数。实际结果显示，瓦斯体积分数实测值与预测值的平均残差为0.019 3，证实了通风数据算
法模型的可靠性，在实现数据高效利用的同时，为矿井进行危险防范提供了重要依据。
关键词： 通风数据算法模型；灰色预测模型；BP神经网络模型；平均残差；危险防范
中图分类号： TD724     文献标识码： A     文章编号： 2095-0802-(2021)06-0133-02
Intelligent Analysis and Research on Mine Ventilation Safety Data
ZHANG Chao
(Donggucheng Coal Industry Co., Ltd. of Zuoyun, Shanxi Coal Imp. & Exp. Group, Zuoyun 037100, Shanxi, China)
Abstract: Based on a large number of ventilation monitoring data of Donggucheng Coal Industry Co., Ltd., this paper established
the ventilation data algorithm model, and forecasted the gas volume fraction through MGM(1,n) grey prediction model and BP neural
network model. The actual results show that the average residual between the measured and predicted gas volume fraction is 0.019 3,
which proves the reliability of the ventilation data algorithm model, and provides important basis for mine hazard prevention while
realizing the efficient utilization of data.

Key words: ventilation data algorithm model; grey prediction model; BP neural network model; average residual; risk prevention