基于大数据的煤矿瓦斯与煤自燃共生灾害风险识别与预警理论研究

The symbiotic disaster caused by gas and coal spontaneous combustion in coal mine is the process and result of multi-factor coupling evolution of complex system. The project aims at the problem of risk identification and early warning of large data with characteristic by massive, dynamic, association, coupling and complex information in the process of the evolution of symbiotic disasters. Safety science, system science, information science, computer science and engineering, cloud computing and other related theories and technologies are used and the combination of theoretical analysis, field data collection and analysis, computer application and laboratory simulation adopted. Characteristic analysis of gas and coal spontaneous combustion will be developed and coupling mechanism of evolution of gas and coal spontaneous combustion will be explored.Risk identification and risk assessment index system of gas and coal spontaneous combustion symbiosis disaster will be built. Disaster information evaluation and screening mechanism with large data characteristics will be sought. Risk identification model of gas and coal spontaneous combustion based on large data will be established. Early warning mechanism of symbiotic disaster will be investigated and dynamic identification and early warning theory and application system of symbiotic disaster risk will be set up. The theory and method of monitoring, identification, inference and early warning with big data technology for preventing and controlling the symbiotic disaster will be explored. It is of great significance and application value to improve the level of prevention and control of gas and coal spontaneous combustion and to ensure the safety production of coal mine.

本项目基于煤矿瓦斯与煤自燃共生灾害是复杂系统多因素耦合演化的过程及结果这一典型特征，针对共生灾害孕育演化过程中“海量、动态、关联、耦合、复杂信息”特征的大数据风险识别与预警过程中急需解决的问题，运用安全科学、系统科学、信息科学、计算机科学与工程、云计算等理论和技术，采用理论分析、现场数据采集分析、计算机应用和实验室模拟研究相结合的方法，开展瓦斯与煤自燃共生灾害特征分析，探究瓦斯与煤自燃共生灾害演化耦合机制，构建瓦斯与煤自燃共生灾害风险辨识与风险判别指标体系，寻求具有大数据特征的灾害信息评估与筛选机制，建立基于大数据的瓦斯与煤自燃共生灾害风险识别模型，研究共生灾害风险预警机制，构建共生灾害风险动态识别与预警理论及应用体系架构，探索实现共生灾害预防与控制的大数据“监测、识别、推理、预警”的理论与方法，对提升瓦斯与煤自燃共生灾害防治水平、保障煤矿安全生产具有重要意义和推广应用价值。

煤炭工业是国家经济快速发展的有力保障，瓦斯、火、水、粉尘等是其开采过程中的主要灾害类型，特别是随着开采深度的增加，瓦斯与煤自燃灾害交织共生，灾害风险不断增大，煤矿安全生产形势愈加严峻。针对共生灾害孕育演化过程中“海量、动态、关联、耦合、复杂信息”特征的大数据风险识别与预警过程中急需解决的问题，运用安全科学、系统科学、信息科学、计算机科学与工程、云计算等理论和技术，采用理论分析、现场数据采集分析、计算机应用和实验室模拟研究相结合的方法，开展了瓦斯与煤自燃共生灾害特征分析，得出了共生灾害是裂隙场、氧气浓度场、瓦斯浓度场与温度场多场时空耦合作用的结果，揭示了瓦斯与煤自燃共生灾害演化耦合机制；基于相似理论构建了三维采空区流场模拟实验系统，开展了共生灾害中诱发因素的互为影响、动态变化和关联特征研究，得出瓦斯运移和分布主要受煤自燃产热影响，瓦斯浓度变化以煤自燃产气效应为主导，在不同高度平面，受产热效应、产气效应和漏风风流影响的比重不同；基于采空区瓦斯特征、采空区漏风条件、采空区蓄热条件、采空区遗煤条件和采空区安全状态等五个方面影响因素，构建了共生灾害风险辨识与风险判别指标体系；基于流场理论与风险识别理论辨识了共生灾害风险的影响因素，通过Pearson系数计算了影响因素与风险的相关度，建立了基于大数据的瓦斯与煤自燃共生灾害风险识别模型，通过将风险评价指标数据化来量化风险影响因素，建立了基于t-SNE+SVR的共生灾害风险预警模型；基于数值模拟结果和预警模型开发了共生灾害预警系统，实现了不同工况下共生灾害的动态演化规律和实时预警。基于时间、瓦斯、氧气、温度等要素，结合煤矿智能化监控系统构建了共生灾害风险动态识别与预警应用体系架构，为实现共生灾害“大数据监测、大数据识别、大数据推理、大数据预警”提供理论和应用基础，提升了煤矿瓦斯与煤自燃共生灾害防治水平。