Research on sensor fault diagnosis method based LVQ neural network and clustering analysis

To meet the robustness of the fault diagnosis algorithm for identifying the novel fault pattern, the method, which combines the supervised classification and unsupervised classification, is proposed in this paper. As the supervised classification, Learning vector quantity neural network is employed to classify sensor mode. As the unsupervised classification, subtractive clustering is applied to identify the novel fault pattern. Finally, the applicability and effectiveness of the proposed methodology is illustrated by flow sensor data of the dynamical system. The result showed that the modal established could meet the robust requirement of fault diagnosis algorithm.