基于规则的航空发动机孔探图像诊断方法研究  

Diagnosis Method Research of Aeroengine Endoscopic Images Based on Rule

作　　者：李华[1] 陈果[1] 陈新波[2] 王洪民 

机构地区：[1]南京航空航天大学民航学院,南京210016 [2]海军航空工程学院青岛校区,山东青岛266000

出　　处：《航空发动机》2015年第3期97-102,共6页Aeroengine

基　　金：国家自然科学基金(61179057)资助

摘　　要：针对军用航空发动机内部损伤频率高、诊断决策时间长的问题,运用基于规则的推理方法,结合案例推理,进行了航空发动机内部损伤维修决策的专家诊断研究,开发了基于孔探图像的航空发动机内部损伤维修决策专家系统。依据领域专家的经验,通过分析得到了对该型发动机的专家知识,并将其转换为if-then的知识规则存放于知识库中;利用图像处理方法,对损伤进行测量,得到损伤的尺寸参数;根据损伤尺寸,依据相应的知识规则,应用正向推理得到损伤的维修决策。运用多个航空发动机内部损伤实例验证了所开发的专家系统的有效性。Aimed at the problem that high-frequency of internal damage and long-time for diagnostic decision making in the military aeroengine,an aeroengine internal damage maintenance decision expert system is developed based on endoscopic image using the Rule Based Reasoning （RBR）,combined with Case Based Reasoning （CBR）.According to the experiences of domain experts,expert knowledge was obtained and saved in the repository after converted to IF-THEN format.Damage size parameters were obtained by using image processing method.The maintenance decision was made through forward chaining method based on the damage size and the corresponding rules.Multiple instances of aeroengine internal damage were used to verify the expert system,and the validity and effectiveness of the results were demonstrated.

关 键 词：叶片损伤 维修决策 规则推理 孔探图像 专家系统 特征提取 航空发动机 

分 类 号：V263.6[航空宇航科学与技术—航空宇航制造工程]