基于云计算的矿井水害智能应急救援方法研究

With the development of cloud computing, Internet of Things, big data and other technologies, smart emergency rescue has become an urgent need and trend of mine production safety. In view of the increase of information sources, the enhancement of numerical simulation ability, and the complexity and difficulty of rescue work in water disaster accidents, this project establishes multi-source heterogeneous and multi-dimensional real-time "one map" through integrating and mining a large amount of data from heterogeneous resources and building an emergency rescue think tank, to promote the sharing of information resources. We will solve the key technologies, such as the information identification and warning for mine water disaster, the numerical simulation of water flow in extractive space and the model of smart emergency evacuation, to promote the fusion of real-time data in the simulation process and improve the timeliness of emergency rescue. The elastic resources for emergency rescue, based on converged infrastructure, unlimited scaling and shared services, can support historical water disaster accident analysis and possible water disaster monitoring and rescue. The ultimate goal of the project is to establish a theoretical method for smart emergency rescue in mine water disaster based on cloud computing, which serves multi-end users and multi-program concurrent computing, supports mine emergency drills and emergency decisions, and provides intelligent services for emergency rescue in mine water disaster to realize the dynamic management and control of prevention, monitoring, early warning, disposal and assessment for water disaster accidents.

随着云计算、物联网、大数据等技术的发展，智能应急救援已成为矿山生产安全的迫切需求和趋势。针对信息源增多、数值模拟计算能力增强、水害事故抢险救援复杂且难度大等情况，本项目通过集成与挖掘来自异构资源的大量数据，构建应急救援智库，建立云计算模式下多源异构、多维度实时“一张图”，推动信息资源共享；解决矿井突(透)水征兆信息识别预警、采掘空间水流数值模拟预测及智能应急疏散模型等关键技术问题，促进实时数据在模拟过程中的融合，提高应急救援时效性；基于融合基础设施、无限扩展和共享服务为应急救援提供弹性资源，支撑历史水害事故分析和可能发生的水害监测救援。项目最终目标将建立基于云计算的矿井水害智能应急救援理论方法，服务于多端用户和多方案并发计算，支持矿山的应急演练和应急决策，为矿井水害应急救援提供智能化服务，实现矿井水害事故预防、监控、预警、处置、评估等全过程动态管理和控制。

随着云计算、物联网、大数据等技术的发展，智能应急救援已成为矿山生产安全的迫切需求和趋势。本项目提出了基于物联网的多源大数据集成与挖掘方法，构建应急救援智库，建立了云计算模式下多源异构、多维度实时“一张图”，推动信息资源共享。提出了利用变权理论、机器学习等方法，实现矿井突(透)水征兆信息捕捉和有效识别预测；基于物联网监测设备的数据变化规律，构建多类型传感器关联预警规则，建立监测设备预警、单一系统预警和综合预警三层次预警体系，增强矿井水害监控预警能力。提出了矿井采掘空间突(透)水蔓延过程数值模拟方法，首次开展了矿井突(透)水蔓延过程物理模拟试验，揭示了水害情况下水流的流动规律；提出了一种基于机器学习的矿井水害灾情研判及预测新方法，实现定量突(透)水模式匹配识别及动态预测矿井水害灾情。提出了时变网络路径规划算法，对逃生人员进行基于水流实时信息的疏散调度，实现智能动态路径并行搜索计算，获得最优应急疏散方案，提高应急救援时效性。提出了涉险人员危险性影响因素的评价指标体系，实现了矿井水害过程中涉险人员危险性的动态评价；首次开展了矿井巷道涉水人员稳定性物理模拟试验研究，验证了水流危险性评价方法的合理性。基于融合基础设施、无限扩展和共享服务为应急救援提供弹性资源，支撑历史水害事故分析和可能发生的水害监测救援。建立了基于云计算的矿井水害智能应急救援理论，服务于多端用户和多方案并发计算，支持矿山的应急演练和应急决策，为矿井水害应急救援提供智能化服务，实现了矿井水害事故预防、监控、预警、处置、评估等全过程动态管理和控制。