Generator insulation fault degradation monitoring method based on partial discharge characteristics

The invention provides a generator insulation fault degradation monitoring method based on partial discharge characteristics. The method comprises the steps of obtaining partial discharge data sets corresponding to various historical insulation faults in a hydro-generator, and building a partial discharge characteristic database according to the partial discharge data sets, wherein the partial discharge data set comprises original data, a characteristic spectrum and a discharge amplitude of partial discharge, training by using the partial discharge characteristic database through a deep learning image recognition algorithm to obtain a fault map recognition model, acquiring a real-time acquisition data set, inputting the data set into the fault map recognition model for characteristic recognition, obtaining various insulation fault data existing in the data set, and calculating corresponding discharge amplitudes, and according to the fault types, adopting a time sequence analysis method to analyze the partial