基于深度神经网络的回采工作面瓦斯涌出量预测  

Prediction of Gas Emission in Mining Face Based on Deep Neural Network

作　　者：宋世伟 张雪 张喜超 景媛媛 Song Shiwei;Zhang Xue;Zhang Xichao;Jing Yuanyuan(Quality Testing Department,Guangli Technology Co.,Ltd.,Zhengzhou Henan 450064,China;School of Mechanical and Electrical Engineering,Zhengzhou Institute of Industrial Application Technology,Zhengzhou Henan 450064,China)

机构地区：[1]光力科技股份有限公司质量测试部,河南郑州450064 [2]郑州工业应用技术学院机电工程学院,河南郑州450064

出　　处：《现代工业经济和信息化》2024年第9期115-116,119,共3页Modern Industrial Economy and Informationization

摘　　要：为了提高煤矿安全性,设计了一种基于深度神经网络的回采工作面瓦斯涌出量预测方法,并在某煤矿回采工作面瓦斯涌出实际量进行了测试分析。研究结果表明:预测结果和实际参数发生了轻微变化,而总体预测结果与瓦斯涌出量变化特点相符,可以较准确反馈涌出量变化幅度。预测误差位于接近0的部位,在-3~3之间,沿两边呈现逐渐下降特点,超过75%的参数预报误差在1.75以内,预测得到误差处于允许范围之内。该研究有助于提高煤矿节能减排的效果,具有很好的实际意义。In order to improve the safety of coal mine,a prediction method of gas emission based on deep neural network is designed,and the actual amount of gas emission in a coal mine is tested and analyzed.The results show that there are slight changes between the predicted results and the actual parameters,and the overall predicted results are consistent with the characteristics of the gas emission quantity,which can accurately feedback the emission quantity variation amplitude.The prediction error is located near 0,between-3 and 3,and presents a gradual decline along both sides.The prediction error of more than 75%parameters is within 1.75,and the prediction error is within the allowable range.This research is helpful to improve the effect of energy saving and emission reduction in coal mines,and has good practical significance.

关 键 词：瓦斯涌出量 回采工作面 深度神经网络 预测误差 

分 类 号：TD712.5[矿业工程—矿井通风与安全]