多源信息融合评价技术是煤层底板突水预测的重要手段，该方法在应用过程中存在专业技术要求高、评价过程复杂繁琐且智能化水平不足的问题。根据突水危险性评价技术流程特点，系统阐述煤层底板突水多源信息综合评价系统的设计原则、设计思路、开发模式和开发方式。构建具有普适性的煤层底板突水评价指标体系、权重计算方法和危险性评价模型等系统核心功能。以GIS技术为基础支撑，依托Visual Studio开发平台，合理调用三方控件(WindowsForm控件、DevExpress控件、ArcEngine控件)，引入随机森林机器学习方法，集成开发从数据处理、存储、权重运算、专题图制作到评价结果可视化输出等多功能一体化的煤层底板突水危险性综合评价系统，实现从源数据到评价结果可视化呈现的“一”系统，使煤层底板突水评价技术更加简便和高效，解决了现有技术依赖多平台计算、评价步骤分离化大且智能化水平低的问题。

The evaluation technology of multi-source information fusion is an important method for the prediction of water inrush from the coal seam floor. However, this method has the deficiencies, such as high professional technical requirements, complicated and tedious evaluation process, and insufficient intelligence during its application. In this study, the design principle, design idea, development mode, and development method of a multi-source information comprehensive evaluation system for water inrush from coal seam bottom were systematically described according to the characteristics of the technical process of water inrush risk evaluation. The key functions of system were built, including the universal evaluation index system of water inrush from coal seam floor, the weight calculation methods and the risk evaluation model. Based on the GIS technology and the Visual Studio development platform, the third-party controls (WindowsForm control, DevExpress control, and ArcEngine control) were reasonably invoked, and a random forest machine learning method was introduced. In this way, a comprehensive evaluation system for the risk of water inrush from coal seam floor was developed, integrating multiple functions from data processing, storage, weighting operation, thematic map production to the visualization output of evaluation results. Thus, the "one" system was implemented from source data to visual presentation of evaluation results, which makes the evaluation technology of water inrush from coal seam floor more simple and efficient, and solves the problems of existing technology relying on multi-platform calculation, large separation of evaluation steps and low intelligent level.