Fault detection and identification: serial form versus simultaneous form

Abstract

AbstractPattern recognition techniques have been widely applied for fault diagnosis. In this paper, two different forms, named serial form and simultaneous form, of pattern recognition based fault detection and identification systems are discussed. The optimal form is selected by a performance assessment rule which is based on the overall fault diagnosis system loss. Beside of considering the fault detection and isolation performance of the classifiers, the misclassification costs in both of the fault detection and isolation stages have also been considered for selecting the optimal form. In order to compare the performance of these two forms, two novel fault detection and isolation approaches integrating kernel principal component analysis (KPCA) and support vector data description (SVDD) are proposed subsequently. A simulation case study of Tennessee Eastman (TE) process is presented to evaluate the proposed fault detection and isolation methods.