【摘要】 露天煤矿的粉尘污染问题已成为制约矿山安全生产的关键因素。随着矿山开采向系统化、复杂化与绿色化方向不断发展，传统粉尘监测手段存在人力成本高昂、数据反馈滞后等明显局限，已难以适配大型露天矿当前的开采需求。为了开发露天煤矿粉尘精准预警技术，解决制约矿山安全生产与生态环境保护的关键问题，分析了哈尔乌素露天煤矿现场监测数据，构建了融合LoRa无线传输与GIS三维可视化的粉尘气象监测平台，实现了PM2.5、PM10、总悬浮颗粒物（Total Suspended Particulates,TSP）等10项参数的实时采集与动态展示，并建立了基于长短期记忆网络（Long short Term Memory,LSTM）的粉尘浓度预报模型。结果表明：实地验证的预测准确率达85%以上，满足矿山粉尘污染“提前预警、精准防控”的工程需求。更多还原

【Abstract】 Dust pollution in open-pit coal mines has become a critical factor constraining mine safety production. As mining operations continue to evolve toward systematization, complexity, and green development, traditional dust monitoring methods face significant limitations such as high labor costs and delayed data feedback, making them increasingly inadequate for the current needs of large-scale open-pit mine. In order to develop precise dust warning technology for open-pit coal mines and solve the key problems that restrict mine safety production and ecological environment protection, the article analyzes the field monitoring data in Ha’erwusu Open-pit Coal Mine, builds a dust meteorological monitoring platform that integrates LoRa wireless transmission and GIS3D visualization, realizes dynamic display of 10 parameters including PM2.5, PM10, Total Suspended Particulates(TSP) and so on,establishes a dust concentration forecasting model based on Long Short Term Memory(LSTM). The results show that the fieldvalidated prediction accuracy exceeds 85%, meeting the engineering requirements for "timely early warning and precise prevention" of mine dust pollution.更多还原