ABSTRACT
A joint on-line fault diagnosis method based on the one-class Support Vector Machine (SVM) and the fuzzy expert system theory was proposed. To the measured data during the on-line faulty diagnosing, first a one-class SVM is applied to judge whether the device is faulty. Then the fuzzy expert system is applied to reason and accurately locate the fault. The result from the conventional fault diagnosing for certain radar indicates that the proposed method has a diagnosis accuracy higher than 90%. The new method has a high fault location accuracy and real-time processing capability, and improves on-line fault diagnosis capabilities.
