摘要
为解决受矿井内粉尘浓度、湿度和温度等现场条件动态变化影响,难以设定最佳降尘参数以满足持续高效降尘的问题。文中将相似性模型试验和多种机器学习算法相结合,构建了一种矿井喷雾降尘动态调控模型,并针对黑龙江双阳煤矿综掘面的产尘问题开展所提模型的实际应用。模型累计考虑20个试验参数,涵盖最佳雾滴粒径参数优选、特征参数对雾化效果影响分析以及降尘效率预测模型3大模块。结果表明:以3个模块所构建的矿井喷雾降尘动态调控模型,可以对综掘面巷道喷雾降尘参数进行实时优化,有效降低现场粉尘浓度,增强喷雾降尘效果,且该方法具有较好的泛化性。
To tackle the discrepancies between onsite dust suppression effects and anticipated dust control outcomes—resulting from dynamic field conditions such as dust concentration,humidity,and temperature—this paper addresses the challenge of establishing optimal dust suppression parameters for sustained and effective control.It integrates similarity model testing with various machine learning algorithms to develop a dynamic regulation model for mine spray dust suppression.The model is applied to address the coal dust control issue at the comprehensive excavation face of the Shuangyang Coal Mine in Heilongjiang.This dynamic regulation model for spray dust suppression takes into account 20 experimental parameters and utilizes 512 sets of experimental data.It encompasses three primary modules:optimal selection of the ideal droplet size parameter,analysis of the impact of characteristic parameters on atomization effects,and a predictive model for dust suppression efficiency.The research findings indicate that the Sauter Mean Diameter(SMD)is the optimal droplet size parameter for characterizing atomization effects in various spray dust suppression models.Regression analysis shows that the droplet size mean diameter(D_(SMD))is directly proportional toα,p_(h),and v,while being inversely proportional to p_(g).Additionally,Q,which is affected by the combined variables of p_(h)and p_(g),exhibits a negative correlation with D_(SMD).Utilizing the Grey Wolf Optimizer(GWO)and Backpropagation(BP)neural network,a dust suppression efficiency prediction model was developed usingα,p_(h),p_(g),Q,v,and D_(SMD)as feature parameters.The model achieved a median relative error of 0.016 and an R^(2)value of 0.954,demonstrating strong predictive accuracy.The dynamic regulation model for mine spray dust suppression,built upon three modules—optimal droplet size selection,analysis of the impact of characteristic parameters on atomization effects,and prediction of dust suppression efficiency—facilitates real-time optimization of spray dust suppression parameters in comprehensive excavation face tunnels.This model effectively reduces onsite dust concentration and enhances dust suppression efficacy,demonstrating strong generalization potential.
作者
刘永立
全金峰
梁云飞
方磊
佟林全
张江石
LIU Yongli;QUAN Jinfeng;LIANG Yunfei;FANG Lei;TONG Linquan;ZHANG Jiangshi(School of Safety Engineering,Heilongjiang University of Science and Technology,Harbin 150000,China;School of Emergency Management and Safety Engineering,China University of Mining and Technology(Beijing),Beijing 100083,China;NHC Key Laboratory for Engineering Control of Dust Hazard,Beijing 102308,China)
出处
《安全与环境学报》
北大核心
2025年第6期2207-2214,共8页
Journal of Safety and Environment
关键词
安全卫生工程技术
机器学习
呼吸性粉尘
降尘处理
参数优化
safety and hygiene engineering technology
machine learning
respirable dust
dust reduction treatment
parameter optimization
