Abstract: In order to realize intelligent decision of drilling engineering risk, considering the characteristic of drilling expert decision making, aiming at the shortcomings of traditional CBR technology, a kind of decision-making model of drilling engineering risk based on ontology and CBR is put forward. Through analyzing risk case knowledge, this paper designs the case structure and builds the case ontology of drilling risk decision-making by the ontology technology, making case representation standardization and semantization, at the same time, providing good scalability and sharing for case knowledge. In order to improve the efficiency of risk case retrieval, it organizes all cases hierarchically according to the risk classification in the case ontology, builds a risk case base. Aiming at the defects of exiting case similarity calculation models, it puts forward a kind of improved semantic similarity calculation model based on ontology, and based on this model, constructs the risk case retrieval model. Through site test, the results show that compared with traditional models, this model can improve the precision ratio and recall ratio of case retrieval effectively. On the basis of above, a prototype system for drilling engineering risk decision is developed, providing more efficient decision support for drilling experts and technicians.

Key words: drilling engineering risk, decision, ontology, case-based reasoning；case retrieval

摘要： 为了实现钻井工程风险智能决策，从钻井专家决策过程的特征出发，针对传统的CBR技术的不足，提出并构建了一种基于本体和CBR的钻井工程风险决策模型。通过对风险决策案例的分析，设计了钻井风险决策案例的结构，并结合本体技术构建了风险决策案例本体，使案例表示更规范化、语义化，同时为案例知识提供了良好的可扩展性与共享性；为了提高案例检索效率，按照本体中风险的类别对案例进行分层组织，建立了风险案例库；针对现有的案例相似度计算模型的缺陷，提出了一种改进的基于本体的语义相似度计算模型，并由此构建了风险案例检索模型，经现场实例测试，结果表明该模型有效地提高了案例检索的查准率和查全率。在以上基础上，开发了钻井工程风险决策原型系统，为钻井专家、技术人员提供了高效的决策支持。

关键词: 钻井工程风险, 决策, 本体, 案例推理, 案例检索