Research on Waveform Identification of Power Fault Recording Based on Instantaneous Characteristic Parameters

When identifying the signal waveform, it is interfered by the clutter in the original power fault record, which leads to the low accuracy of the identification result. Therefore, the research of power fault record waveform identification based on instantaneous characteristic parameters is proposed. A multicomponent real-valued power fault recorded signal is decomposed in a non-recursive way, which is transformed into a discrete quasi-orthogonal limited bandwidth sub-signal set, and the effective bandwidth is extracted as the target instantaneous characteristic parameter. In the phase of waveform identification, the unsupervised adaptive framework of countermeasure domain (FA-AUDA) is applied, and the countermeasure network (CGAN) is generated with the help of specific conditions (power failure), so as to enhance the instantaneous characteristic parameters of source domain data in the feature space and realize waveform identification. In the test results, the design method has the highest accuracy for the identification of power fault waveform in different crossing States, and it is always stable above 0.70.