摘要
应用基于主元分析的故障诊断方法对浮式油轮生产储油卸油系统(FPSO)进行故障检测与诊断研究。选取FPSO油气水分离系统的18个主要过程监控变量为研究对象,通过对系统历史数据进行预处理分析,建立主元模型;利用主元模型对仿真实时数据进行故障检测,应用SPE统计法和Hotelling统计法判断系统是否发生故障;使用贡献图法实现故障分离。研究结果表明:基于主元分析的故障诊断方法可以准确地对FPSO生产过程的早期故障进行检测和诊断;且对于系统的细小扰动,动态主元分析法的故障诊断能力优于主元分析法。
Used fault diagnosis method based on Principal Components Analysis（PCA） to research the fault monitoring and diagnosis of Floating Production Storage and Off-loading system（FPSO）. Selected 18 monitor variables in FPSO oil process as research object to build up the principal components model via pretreated history data of normal operation condition. The principal components model was applied to research system fault monitoring and diagnosis ,judged the SPE statistic and Hotelling statistic to monitor the fault. The principal components contribution diagram was used to diagnose the fault source. The result shows that, the fault diagnose method based on PCA can realized FPSO earlier period fault monitoring and diagnosis accurately, and the capability of PCA fault diagnosis is better than the Dynamic Principal Components Analysis （DPCA）for the small disturbance.
出处
《化工自动化及仪表》
CAS
2008年第4期7-11,共5页
Control and Instruments in Chemical Industry
基金
天津市高等学校科技发展基金资助项目(2006BA22)
