顶板岩层结构劣化是引起锚杆支护煤巷冒顶的重要因素之一，顶板失稳与其组合结构有关，不同组合结构的顶板在锚固孔施工时具有不同的动力响应特性。因此，在钻进顶板锚固孔时动态识别其内部特征，找到岩层结构劣化区域，采取合理的锚杆支护参数是预防冒顶的有效方法。项目采用冲击（钻削）破岩理论和动力分析模拟方法，确定判断岩层类型和厚度的锚固孔钻进动力响应敏感指标。利用相关传感器获取实验室相似模拟方案和井下钻进实验的顶板锚固孔钻进动力响应指标，并与钻取岩芯对比，综合分析形成敏感指标库。以此为数据源，利用模式识别方法，建立煤巷顶板岩层类型识别模型，得出锚杆支护煤巷冒顶隐患快速智能识别技术与方法并通过实践验证。通过项目的研究，可解决该领域煤巷掘进过程中顶板岩层结构特征难于实时动态弄清的瓶颈问题，促进煤巷锚杆支护参数设计方法的更新，同时对于预防锚杆支护煤巷冒顶事故的发生具有重要的意义。

Deterioration of the rock structure is one of the most major factors to make the bolt support caving in the roadway roof. Instability of the roof is related to its composite structure. The roof with different composite structure usually presents distinct dynamic response characteristic under construction of anchoring. The effective method of preventing caving is to recognize dynamically internal characteristic of anchoring with drilling and search the deteriorating zone of rock structure, as well as put forward to reasonable supporting parameter of the bolt. Impacting (drilling) breaking rock theory and method of dynamic analysis and simulation has been introduced to decide drilling dynamics sensitive index of the anchor hole with different rock type and depth. The relevant sensor has been exploited to gain the similar simulating program and response index of anchor hole drilling dynamics in the roof under the roadway, which have been compared with drilling core and analyzed integrated to make the sensitive index database. Depended on the database and utilized the pattern recognition method, the recognition model of rock type in the roof has been presented. Furthermore, the fast intelligent recognition technology of potential caving accident in the bolt supporting roadway has been verified through practice. Once the research is implemented, the chock point that is hard to dynamically find rock structure characteristic out during drilling will be solved and the design method of bolt supporting parameter will be developed, as well as the potential caving accident will be prevented effectively.