Time-Frequency Based Feature Extraction and Classification for Fault Diagnosis in Electric Drives

The detection of non-catastrophic faults can be used to determine the remaining life of an electric drive. As the frequency and severity of these faults increase, the working life of the drive decreases, leading eventually to failure. In recent years, various signal processing techniques have been proposed to extract useful features and to classify the fault signals. Previous work has focused on the steady-state behavior of faulted drives. In this paper, we are interested in detecting the transient phenomena associated with the inception and clearing of the faults. We propose time-frequency representation based methods for extracting information about the transient phenomena observed for electrical and mechanical faults. An efficient feature extraction method and a corresponding classification algorithm are proposed for real-time detection and classification of faults.