Fault diagnosis of rocket engine ground testing bed based on KPCA and SVM

To solve the problem of fault diagnosis for liquid propellant rocket engine ground testing bed,a fault diagnosis approach based on kernel principal component analysis(KPCA) feature extraction and support vector machines(SVM) multi-classification is proposed.After extracting the feature of testing bed standard fault samples,a hierarchical support vector machine(H-SVM) for multi-classification is established and trained through feature analysis.Then,by projecting the testing data to the principal component,the fault status is identified using the trained multi-classifier.Through exhausting the capability of non-linear feature extraction of KPCA and the small sample generalization of SVM,this approach resolves the pattern recognition problem of small sample and nonlinearity in fault diagnosis of testing bed.The experimental results indicate that this approach is available and effective.