基于实例和规则集成推理的钻井事故诊断处理系统

石油钻井中的事故诊断与处理既凭经验又离不开相关知识，为此，将基于实例推理（CBR）和基于规则推理（RBR）两种方法集成为一体，以更好地解决该类问题。介绍了CBR和RBR混合推理的钻井事故诊断与处理专家系统。该系统根据钻井事故诊断与处理的特点，将专家经验采用产生式规则表示，并建立基于规则的知识库；将实例采用规则、框架和过程的混合知识表达形式表示，并建立系统的实例库。实例库包括用RBR系统自动生成的基本实例库及无确定规则的特殊实例，在使用中还可不断地增加新实例以提高系统判断复杂事故的能力。该系统采用RBR确定事故类型，再利用CBR获取事故处理措施。由于采用了混合推理，事故诊断快速可靠，自学习能力也得到了很大提高。

A Drilling Accident Diagnosis and Treatment System Based on Case-Based Reasoning and Rule-Based Reasoning

The diagnosis and treatment of drilling accidents are based on experience and related knowledge,therefore,the combination of case based reasoning(CBR) and rule based reasoning(RBR) will solve the drilling accidents better.This paper presents an expert system for diagnosis and treatment of drilling accidents.According to the characteristics of drilling accident diagnosis and treatment,expert experiential knowledge was expressed by productive rules and established the rule-based knowledge database,and the cases were expressed by rule,frame and process mixed representation method and established the case-based case database.Case database includes base cases generated automatically by RBR system and undetermined special cases,and the new cases cane be incorporated into the case database to improve the ability in diagnosing the complex accidents.This system uses RBR to determine the type of the accidents,and the CBR to obtain the treatment.Because of the mixed reasoning,the diagnosis of the accident is reliable and quicker and the self-learning is greatly improved.