Research on Fault Diagnosis of Tennessee Eastman Process Based on KPCA and SVM

Abstract

Based on principal component analysis (PCA) and support vector machine (SVM), a new method for the fault diagnosis of TE Process is proposed. The fault recognition based on kernel principal component analysis (KPCA) is analyzed and SVM is employed as a classifier for fault classification. To establish a more efficient SVM model, genetic algorithm (GA) is used to determine the optimal kernel parameter ? and penalty parameter C of SVM with the highest accuracy and generalization ability. The classification accuracy of this GA-SVM approach is tested by real data of TE Process and compared with some other related methods such as artificial neural network. The experimental results indicate that the classification accuracy of this GA-SVM is more superior than that of some artificial neural network.