基于本体和信号分析的汽轮发电机组故障诊断方法研究  

A research on fault diagnosis method of turbine generator sets based on ontology and signal analysis

作　　者：艾科勇 张永明 王文斌 栗宇 剡昌锋[1] Ai Keyong;Zhang Yongming;Wang Wenbin;Li Yu;Yan Changfeng(School of Mechanical and Electrical Engineering,Lanzhou University of Technology, Gansu Lanzhou, 730050, China;Gansu Luqiao Feiyu Transportation Facilities Co., Ltd., Gansu Lanzhou, 730050, China;Daqin Railway Co., Ltd., Shanxi Taiyuan, 030045, China)

机构地区：[1]兰州理工大学机电工程学院,甘肃兰州730050 [2]甘肃路桥飞宇交通设施有限责任公司,甘肃兰州730050 [3]大秦铁路股份有限公司,山西太原030045

出　　处：《机械设计与制造工程》2021年第11期65-70,共6页Machine Design and Manufacturing Engineering

基　　金：国家自然科学基金资助项目(51765034,51165018)。

摘　　要：针对传统的单一故障诊断方法在汽轮发电机组故障诊断中具有一定的局限性,研究了本体和信号分析在故障诊断领域的优势,提出一种适用于汽轮发电机组故障诊断的方法。借助Protégé4.3构建汽轮发电机组故障诊断本体知识库,基于EEMD、排列熵和SVM的信号分析方法辨识出故障类型;设计一种语义映射方法,将信号分析的辨识结果与本体知识库中的实例相关联,并对关联得到的本体实例进行推理,进而得到故障原因和维修策略。最后针对转子常见故障进行实例测试,结果表明该方法能够有效诊断汽轮发电机组常见故障,且诊断结果全面、准确。Aiming at the limitation of the traditional single fault diagnosis method in the fault diagnosis of turbine generator sets,it studies the advantages of ontology and signal analysis in the field of fault diagnosis,and puts forward a method suitable for the fault diagnosis of turbine generator sets.With the help of Protégé4.3,the fault diagnosis ontology knowledge base of turbine generator sets is constructed,and the fault types are identified based on the signal analysis methods of EEMD,permutation entropy and SVM.A semantic mapping method is designed to associate the identification results of signal analysis with the cases in the ontology knowledge base,and the associated ontology cases are reasoned,then the fault causes and maintenance strategies are obtained.Finally,the common faults of the rotor are tested.The results show that the method can effectively diagnose the common faults of the turbine generator sets,and the diagnosis results are comprehensive and accurate.

关 键 词：汽轮发电机组 故障诊断 本体推理 信号分析 语义映射 

分 类 号：TH17[机械工程—机械制造及自动化] TP182[自动化与计算机技术—控制理论与控制工程]