Identification of immune models for fault detection

Abstract

In this paper, a novel approach to model-based fault detection for non-linear systems is presented. An immune model of the system is used for the generation of residual. The orthogonal least-squares method is implemented to select the significant receptor vectors of the immune model. After the model identification, the filtered residual scheme and the fault alarm concentration are applied for the fault detection. To verify and demonstrate the performance of the proposed methodology, a simulation example on a two-link robot was studied. The results show the effectiveness and robustness in both system identification and fault detection.