我国煤矿事故有96.5%是由人为因素导致的，其中由于违章作业而引发的人因事故高达70-80%。在此背景下，为有效控制违章行为的发生频次及概率、实现煤矿企业违章信息高效化管理、营造规范的作业环境，本文借助物联网技术开发矿工违章行为识别与控制系统以实现对矿工违章行为的监测、识别和控制，辅助管理人员高效管理违章行为，提升矿井安全生产水平和事故预防能力。 首先，总结分析违章行为的概念、分类及致因机理，明确煤炭行业违章识别与控制研究方面的欠缺；依据系统预设效果，对系统设计中涉及的关键技术进行阐述分析，明确本系统设计中信息采集设备及信息加工处理模式，从多角度对系统的可行性进行论证分析；通过分析矿工违章行为的发生机理，借助CBR搭建矿工违章行为的框架结构，确定回采工作面各工种违章行为的识别规则，构建基于CBR的矿工违章行为的“人工+智能”的双重违章识别与控制模型，基于此模型对矿工违章行为的识别过程和控制过程进行了详细的分析。 然后，明确系统设计原则和设计方案，对矿井井下和地面的网络结构进行了设计。满足各项基本要求之后，将系统的功能模块化为主要功能模块、辅助功能模块和其他功能模块（系统配套功能）三大板块，数据库是信息系统的核心模块，故将此模块进行了单独设计。依托Visio软件对系统各个界面及窗口进行初步设计，借助Eclipse和Java等开发软件对各个界面进行了实现，在HTML、CSS、JavaScript、Eclipse、Tomcat以及MySQL等多软件开发平台的支撑下实现了系统的各项功能。 最后，运用LoadRunner性能测试软件和黑盒测试软件分别对系统的性能和功能进行了检测，结果表明：系统不仅能够满足150个用户同时登录系统并进行数据请求的操作，而且能够实现各个功能模块的实际功能，充分表明了矿工违章识别与控制系统的可操作性和实用性。该系统将实现违章处理模式由“事后控制”向“事中控制”的转变，对构建安全矿山和提升矿山安全管理水平具有重要意义。

In our country, nearly 96.5% of coal mine accidents are caused by human factors, 70-80% of which are caused by violations. In current production situation, in order to effectively reduce the frequency and probability of violations, realize the efficiency of violative information management, and create a standardized working environment, the paper have developed an identification and control system for miners' violations by the Internet of Things technology in order to monitor, identificate and control miners' violations, assist managemers to control violations effectively, and improve mine safety production and the capabilities of accident prevention. Firstly, the study summarized and analyzed the concept, classification and causal mechanism of violations, and clarified the lack of current research on the identification and control system of violations in the coal industry. The key technologies involved in the system design are expounded according to the preset effect of the system. The information acquisition equipment and information processing mode in the design of the system are clarified, and the feasibility of the system was analysed from multiple angles. By analyzing the mechanism of miners' violations, the framework of miners' violations was constructed by CBR. The rules for identificating various types of violations in the mining face were formulated, and a double miners' violations’ identification and control model based on the "artificial and intelligence" of CBR was proposed, and then the process of identification and control of miners' violations are analyzed in detail. Then, the principles and schemes of system design were clarified, the ground network structure was designed. After satisfied the basic requirements, the system functions were divided into three parts: main functional module, auxiliary function module and other functional module (system supporting function), the database is the core module of the information system, so it was designed separately. Designing the various interface and the window of system by Visio, using Eclipse and Java development software to implemented each interface, accomplished system’s various functions under the support of HTML, CSS, JavaScript, Eclipse, Tomcat and others. Finally, the performance and function of the system were tested by LoadRunner performance test software and black box function test software. The results showed that the system could meet 150 users to log in to the system and perform data request operations and realize the system’s various functions. The actual function fully demonstrates the operability and practicability of the identification and control system.The system will realize the transition from the "post-event control" to the "in-process control" of the violations’ processing mode, which is of great significance for building a safty mine and improving mine safety management level.