Fault Diagnosis Based on Artificial Immune and Principal Component Analysis

On the basis of analyzing on the comparability of the recognizing principle between antigens and antibodies, mathematical model and Singular Value Decomposition (SVD) of matrix, an approach to fault diagnosis combining Principal Component Analysis (PCA) with Artificial Immune is proposed. PCA is used to abstract the characteristic and reduce the dimensionality of the data. SVD is used to get the pairs of antibody and antigen of data to diagnose faults. The data from Tennessee Eastman (TE) process simulator is used to evaluate the effectiveness of this approach. Simulation result shows that the average fault diagnosis ratio for TE process can be up to 84% and this approach is practicable.