摘要
煤矿发生瓦斯灾害前,往往伴随瓦斯浓度异常,准确判断瓦斯浓度是进行瓦斯突出预测、通风设计等工作的基础。通过机器学习的方法,将影响瓦斯浓度的多种因素进行综合计算,探索一种基于机器学习的多因素煤矿瓦斯浓度预测技术,通过构建模型、训练模型、实际使用,计算得出预测值与实际值的误差,并结合实际生产矿井数据,进行验证计算,验证此种方法的可行性,将瓦斯预测技术由被动式变为主动式,为煤矿瓦斯浓度预测提供新思路,同时将大数据、深度学习等智能化技术引入至煤矿瓦斯治理中,具有广阔的应用前景。
Before a gas disaster occurs in a coal mine,it is often accompanied by abnormal gas concentration.Accurately determining gas concentration is the basis for gas outburst prediction,ventilation design,and other related work.We use machine learning method to comprehensively calculate various factors that affect gas concentration,and explore a multi-factor coal mine gas concentration prediction technology based on machine learning.By constructing a model,training the model,and using it in practice,the error between the predicted and the actual value is calculated,and the feasibility of this method verified through actual production mine data.The gas prediction technology is transformed from passive to active,providing a new approach for predicting coal mine gas concentration.Introducing intelligent technologies such as big data and deep learning into coal mine gas management is of widely practical prospect.
作者
徐平安
张若楠
周小雨
赵琦琦
XU Ping’an;ZHANG Ruonan;ZHOU Xiaoyu;ZHAO Qiqi(Ping’an Coal Mining Engineering Technology Research Institute Co.,Ltd.,Huainan 232000,China)
出处
《陕西煤炭》
2024年第3期88-91,144,共5页
Shaanxi Coal
关键词
机器学习
煤矿瓦斯
线性回归算法
浓度预测
machine learning
coal mine gas
linear regression algorithm
concentration prediction
