基于PCA和C-SVM的涡轮部件故障诊断

针对某型三轴燃气轮机高、低压涡轮部件容易出现的8种故障,提出一种基于PCA(主成分分析)与C-SVM(C-支持向量机)相结合的涡轮部件故障诊断模型.采用主成分分析方法对表征涡轮部件故障模式的测量参数进行特征提取,选择对故障模式影响最大的若干主成分作为C-SVM的输入样本,进而对高、低压涡轮部件故障进行诊断.通过实验表明,即使在较少样本的情况下,应用PCA与C-SVM相结合仍能取得较好效果.

Fault diagnosis of turbine based on principal component analysis and C-SVM

The model of turbine component’s fault diagnosis is proposed based on PCA and C-SVM according to the 8 kinds of fault form high-pressure turbine and low-pressure turbine of triple-axial gas turbine.Principal component analysis is being used to obtain feature extraction to the measuring parameters that to express turbine component failure mode.The principal components are chosen which have an main effect on fault pattern as input samples of C-SVM to detect the fault of high-pressure turbine and low-pressure turbine.The results show that the good effects can be achieved by using principal component analysis and C-SVM even though has less samples.