我国95%的煤矿是井工开采，受煤层地质赋存条件等客观因素的制约，煤矿各种灾害严重。瓦斯灾害始终是煤矿安全生产的最大威胁，重、特大瓦斯灾害(煤与瓦斯突出、瓦斯爆炸)事故时有发生，严重影响煤矿职工的生命安全和煤炭工业的可持续发展。 智能决策支持系统(IDSS)是在决策支持系统(DSS)的基础上集成人工智能的专家系统(ES)而形成的，是管理决策科学、运筹学、计算机科学与人工智能相结合的产物，它利用专家系统(ES)技术，预先把专家(决策者)的建模经验整理成计算机表示的知识，组织在知识库中，并用称为推理机的一组程序来模拟决策专家的思维推理，形成一个智能的部件，适用于处理半结构化和非结构化的问题。 本文充分利用煤矿原有的资料和记录，建立以数据库、模型库和知识库为主体的智能决策支持系统。其中数据库存储初始数据和中间数据，包括煤矿近年来的安全检查记录、隐患排查记录以及历年瓦斯事故记录等；模型库存储对问题进行建模产生的各种模型的参数和处理问题的不同方法，重点介绍了模糊事故树模型的原理和建立；知识库存储先验知识和处理问题过程中获取的知识，包括煤矿安全规程和事故分析报告等。通过问题处理系统，对煤矿瓦斯危险源进行智能分析和评价，为煤矿瓦斯危险源辨识工作提供了相关的辅助决策信息，为决策者制定方案、采取相应的防范措施提供了方法。

95% of the coal mines in China are underground mines, which are suffering from all kinds of severe disasters because of the restrictions of the objective factors such as the geological deposit condition of coal seam. Among all of these disasters, firedamp calamity is the biggest threat to the safety in coal production, while heavy or even extraordinary serious gas accidents (coal or firedamp outbursts, gas explosion) arise frequently, which has badly influenced the safety of the employees and the sustainable development of the coal industry. Based on the Decision Support System(DSS), Intelligent Decision Support System(IDSS) integrates Expert System(ES) of Artificial Intelligence, it’s the offspring of Management Decision Science, Operational Research, Computer Science and Artificial Intelligence. It used the technology of Expert System, converted model building experience of the experts into the knowledge which expressed by computer in the Knowledge Database in advance, and used a set of procedures called Inference Engine to simulate the thought and inference of the experts, to form an intelligent component. It can be applied to half-structured and un-structured problems. This paper took good advantage of data and records in mine, established the IDSS which included Database, Model Database and Knowledge Database. The Database stored initial data and mid data, included recent safety inspection records, hidden trouble control records, gas accident records in mine and so on. The Model Database stored all kinds of model parameters from modeling and different methods of solving the problem, introduced the principle and building of fuzzy fault tree model primarily. The Knowledge Database stored prior knowledge and knowledge from solving the problem, included Mine Safety Regulation, accident analysis report and so on. Based on the problems processing system, IDSS could analyze and evaluate the mine gas dangerous source, provide interrelated auxiliary decision-making information for the work of mine gas dangerous source identification, and provide methods for framing plans and carrying out corresponding measures.