基于专家系统的开发动态异常井诊断模型

Development of Dynamic Abnormal Well Diagnosis Model Based on Expert System

油井异常的影响因子复杂多变又相互制约,为了提高异常定位的准确性设计了基于专家系统的开发动态异常井诊断模型(DDAWD)。模型建立数据层-知识层-规则层(DKR)三层知识结构,该知识表示方式不仅保留产生式的结构而且与自然语言构成映射。基于DKR知识结构推理采用前后件推理方式并通过反射执行调用匹配函数来完成对问题井的异常定位。通过对某采油厂的生产数据分析得出实验结果并证明DDAWD模型提高了诊断结果的准确率。

The factor of oil well abnormal are complicated and restricted with each other. In order to improve the accu- racy of abnormal well diagnosis decreases, development dynamic abnormal well diagnosis model（DDAWD） is designed. The model establishes three layer structure of knowledge which contains Knowledge Layer, Rule Layer, and Data Layer（DKR）. The knowledge expression method not only retains the structure of production but also forms a mapping with natural lain guage. Based on the DKR knowledge expression, the production inference is used for reasoning process, and the position of abnormal wells is finished through reflection execute matching function. Test result through analyzing of the production data of a certain oil production plant proves that the model improves the accuracy of diagnosis results.