工矿自动化2026，Vol.52Issue(4)：45-54,87,11.DOI:10.13272/j.issn.1671-251x.18271
基于边缘计算架构与流体运移机理的煤矿瓦斯主动防控系统
Active gas prevention and control system in coal mines based on edge computing architecture and fluid migration mechanism
摘要
Abstract
To address the nonlinearity and time variability of gas accumulation in underground coal mines and the control lag of traditional centralized monitoring,this paper proposed a design scheme for an intelligent active gas prevention and control system in coal mines based on edge computing architecture and fluid migration mechanism.The system adopted a master-slave collaborative star network topology and logically divided the prevention and control tasks into two layers,namely the sensing edge and the decision-making edge.The sensing edge completed oversampling and digital filtering of gas sensor signals at the source end and directly output high-confidence gas concentration values,thereby eliminating interference noise caused by long-distance transmission of analog signals at the physical layer.In coordination with the Time Division Multiple Access(TDMA)communication protocol,the sensing edge transmitted valid data only within the allocated time-slot window,which significantly reduced the probability of channel congestion and the overall system power consumption.Unlike traditional rigid threshold-based on-off control,the decision-making edge was embedded with a fuzzy PID algorithm based on the gas migration mechanism.When an abnormal deviation in gas concentration or an abnormal rate of change was detected,the system predicted the trend of gas accumulation and dynamically adjusted the ventilation fan speed through high-frequency pulse-width modulation signals,thereby avoiding the uncertainty caused by long-distance communication.The simulation and prototype test results showed that,under a harsh radio-frequency environment with a 20％packet error rate,the deterministic interaction delay of control instructions was minimized to 0.5 s.Under the conditions of±20％model mismatch in flow-field parameters and sudden gas outburst,the algorithm showed very strong convergence robustness,and the dynamic overshoot was strictly limited to within 3.4％.The introduced safety-constraint priority mechanism shortened the full-link physical safety disposal time from 72.3 s under the conventional strategy to 41.7 s,while effectively avoiding mechanical fatigue damage to the ventilation fan and achieving rapid active defense against the risk of gas accumulation and smooth recovery.关键词
瓦斯主动防控/边缘计算/强电磁干扰/模糊PID/流体运移机理/TDMA/感知边缘/决策边缘Key words
active gas prevention and control/edge computing/strong electromagnetic interference/fuzzy PID/fluid migration mechanism/TDMA/sensing edge/decision-making edge分类
矿业与冶金引用本文复制引用
张科学,李伟涛,许雯,吕鑫淼,刘晨阳,李成章,陈学习,王晓玲,马砺,张美长,王猛,李小磊,闫星辰..基于边缘计算架构与流体运移机理的煤矿瓦斯主动防控系统[J].工矿自动化,2026,52(4):45-54,87,11.基金项目
国家自然科学基金资助项目(51804160) （51804160）
深部岩土力学与地下工程国家重点实验室(北京)开放基金资助项目(SKLGDUEK1822) （北京）
中国科协科技智库青年人才计划项目(20220615ZZ07110397) （20220615ZZ07110397）
中央高校基本科研业务费资助项目(3142021007,3142019009). （3142021007,3142019009）
