基于模糊贝叶斯网络的汽轮机组故障诊断研究  

Research on Fault Diagnosis of Turbo-Generator Set Based on Fuzzy Bayesian Network

作　　者：张栋良 汪刘峰 洪勤勤 张凯文 ZHANG Dong-liang;WANG Liu-feng;HONG Qin-qin;HANG Kai-wen(College of Automation Engineering,Shanghai University of Electric Power,Shanghai200090,China)

机构地区：[1]上海电力大学自动化工程学院,上海200090

出　　处：《计算机仿真》2022年第7期476-481,共6页Computer Simulation

基　　金：国家自然科学基金青年科学基金(51906133);上海市科委地方能力建设项目(18020500900)。

摘　　要：针对汽轮发电机组结构高度耦合,故障诊断知识复杂,故障诊断过程需要不确定性推理的问题,提出了一种本体、模糊和贝叶斯网络(BN)三者结合的故障诊断方法。该方法首先利用本体理论建立汽轮发电机组故障诊断本体库,并通过建立规则将本体模型转化成BN结构;然后将Leaky Noisy-Or模型和模糊综合评价法结合获取条件概率表(CPT);最后,结合BN结构模型和CPT得到完整的汽轮发电机组诊断模型。实例验证结果表明,该方法应用在汽轮发电机组故障诊断中切实可行,具有较高的准确性。Aiming at the problems that the structure of turbo-generator set is highly coupled,the fault diagnosis knowledge is complicated,and the fault diagnosis process requires uncertainty reasoning,a fault diagnosis method combining ontology,fuzzy and Bayesian network(BN) is proposed.In this method,firstly,the ontology database of fault diagnosis of steam turbine generator set was established by using ontology theory,and the ontology model was transformed into BN structure by establishing rules;Then the Leaky Noisy-Or model and the fuzzy comprehensive evaluation method were combined to obtain the conditional probability table(CPT);Finally,combining BN structure model and CPT were combined to get a complete diagnosis model of turbo-generator set.The example verification results show that the application of this method in the fault diagnosis of turbo-generator set is feasible and has high accuracy.

关 键 词：汽轮发电机组 故障诊断 贝叶斯网络 本体 模糊集理论 

分 类 号：TK268[动力工程及工程热物理—动力机械及工程] TP181[自动化与计算机技术—控制理论与控制工程]