基于粗糙集理论的航空发动机滑油光谱诊断专家系统知识获取方法研究

Knowledge Acquisistion of Aero-engine Spectrometric Oil Diagnosis Expert System Using Rough Set Theory

针对航空发动机滑油光谱诊断专家系统的知识获取问题,本文建立了基于粗糙集理论的航空发动机滑油光谱诊断专家系统知识获取模型。首先建立反映光谱元素浓度及元浓度梯度与发动机磨损故障之间关系的典型故障样本集;然后运用粗糙集理论的离散、约简及规则提取等算法,从大量的故障样本中自动获取知识规则,并将知识规则存储于专家系统知识库中;最后,建立推理机,运用一定的推理策略实现发动机磨损故障诊断。本文利用航空发动机实测的油样光谱数据对所建立的粗糙集知识获取方法,进行了实例验证。

The paper studies the knowledge acquisition of aero-engine spectrometric oil diagnosis expert system and establishes the knowledge acquisistion model with the rough set theory. Firstly, it establishes the representative sample sets which reflect the relationship between the spectrum element content and spectrum element content gradient and the wear faults of an aero-engine. Secondly, it uses the discretization, reduction and rule extraction algorithms of the rough set theory to acquire automatically knowledge rules from abundant fault samples, and saves the rules into the knowledge base of the expert system. Finally, it establishes a reasoning machine and uses some reasoning strategies to carry out the wear fault diagnosis of the aero-engine. With aero-engine spectrometric oil data, the paper verifies the rough set＇s knowledge acquisition method.