摘要
为了解决煤矿瓦斯异常涌出预警难题,基于瓦斯涌出数据动态波动特征确定异常指标,通过神经网络数据挖掘建立瓦斯异常涌出预警模型,以“感知-物联-支撑-应用”为研究架构,以“数据采集-动态分析-实时预警-统计分析”为研究思路,构建了一套瓦斯异常涌出预警平台,实现了瓦斯浓度变化监测点异常回溯分析与异常涌出预警报警等功能。系统在陕煤集团神木红柳林矿业有限公司得以应用,满足了《国家智能化煤矿验收办法》中关于瓦斯灾害智能化建设要求,缩短了红柳林煤矿瓦斯防治关键信息响应时间,提升了部门瓦斯防治工作效率与管理水平。
In order to solve the problem of abnormal gas outflow early warning in coal mines,based on the dynamic fluctuation characteristics of gas gushing data,the anomaly index is determined,and the gas abnormal gushing early warning model is established through neural network data mining,with"perception-IoT-support-application"as the research framework,and"data collection-dynamic analysis-real-time early warning-statistical analysis"as the research idea,a set of gas abnormal gushing early warning platform is constructed,which realizes the functions of abnormal back-trace analysis of gas concentration change monitoring point and abnormal gushing early warning alarm.This system was applied in Shaanxi Coal Group Shenmu Hongliulin Mining Co.,Ltd.,which met the requirements for intelligent construction of gas disasters in the National Intelligent Coal Mine Acceptance Measures,reduced the response time of key information on gas prevention and control in Hongliulin coal mine,and improved the efficiency and management level of gas prevention and control in the department.
作者
崔聪
舒龙勇
梁银泉
杨建
刘正帅
朱南南
CUI Cong;SHU Longyong;LIANG Yinquan;YANG Jian;LIU Zhengshuai;ZHU Nannan(China Coal Research Institute,Beijing 100013,China;Yunnan Dongyuan Zhenxiong Coal Industry Co.,Ltd.,Zhenxiong 657204,China)
出处
《煤炭技术》
CAS
2024年第3期160-164,共5页
Coal Technology
基金
煤科院科技发展基金项目(2021CX-Ⅱ-10)。
关键词
瓦斯异常涌出
智能化
实时预警
模型指标
系统研发
gas abnormal outflow
intellectualization
real time early warning
model indicators
system development
