目前煤矿火灾监测系统实现了对矿井煤自燃标志性气体、温度、烟雾、火焰等部分指标的单独监测，未对煤矿火灾相关因素进行有效、全面、统一的监测。针对该问题，从内因、外因2个方面分析了煤矿火灾潜在危险因素，提出一种分源分区监测火情态势的方法。内因火灾方面，主要针对较易发生火灾的工作面采空区、密闭采空区及人工自然发火观测点等进行监测；外因火灾方面，主要针对机电硐室及其配电点、带式输送机系统、电缆等方面进行监测。建立了煤矿火灾分源分区监测指标体系，采用人工监测或在线监测的方式定期采集或更新火灾特征参量数据，按数据采集方式及影响程度，将火灾监测指标分为动态指标、静态指标和关联指标。设计了火灾智能预警系统的总体架构和业务流程，采用基于多指标联合逻辑推理的预警方法实现内因火灾预警，采用基于D−S 证据理论的多参量融合预警方法实现外因火灾预警。现场试验结果表明，火灾智能预警系统实现了对矿井火灾的有效监测预警，具有煤矿火灾风险预警“一张图”可视化展示功能，同时具备火灾智能模拟演示功能及避灾路线动态规划功能。

Currently, the coal mine fire monitoring system has achieved separate monitoring of some indicators such as the iconic gases, temperature, smoke, and flame of coal spontaneous combustion in mines.But the system has not effectively, comprehensively, and uniformly monitored the factors related to coal mine fires. In order to solve this problem, potential risk factors of coal mine fires are analyzed from two aspects: internal and external factors. A method of monitoring fire situation in different sources and areas is proposed. In terms of internal fires, monitoring is mainly carried out on goaf areas, enclosed goaf areas, and artificial natural fire observation points that are prone to fires. In terms of external fires, monitoring is mainly carried out on the mechanical and electrical chambers and their distribution points, belt conveyor systems, cables, and other aspects. A monitoring index system for coal mine fire sources and areas has been established. The system regularly collects or updates fire feature parameter data through manual or online monitoring. According to the data collection method and impact degree, fire monitoring indicators are divided into dynamic indicators, static indicators, and related indicators. The overall architecture and business process of a fire intelligent warning system is designed. The system uses a warning method based on multi index joint logical reasoning to achieve internal fire warning, and uses a multi parameter fusion warning method based on D-S evidence theory to achieve external fire warning. The on-site test results show that the fire intelligent warning system has achieved effective monitoring and warning of mine fires, with a visual display function of a coal mine fire risk warning "one picture". The system has a fire intelligent simulation demonstration function and a dynamic planning function for disaster avoidance routes.