Expansion and Identification of High Voltage Cable Joint Defect Partial Discharge Data

With the clear goal of “peak carbon dioxide emissions and carbon neutrality”, many wind and solar farms in my country have used aluminum cables as directly buried transmission cables to reduce costs and utilize clean energy. As a high-voltage cable section, the intermediate joint of the cable is the weakest link in the cable line, and its electrical stress is relatively concentrated and prone to failure. The partial discharge pattern recognition of the cable can accurately evaluate its state, but at present, due to the limitation of the partial discharge detection system conditions, the collected partial discharge samples have problems such as lack of data and uneven distribution, which leads to the phenomenon of poor generalization ability of the network model. In order to solve the above problems, this paper selects the intermediate joint of 10 kV XLPE cable, which accounts for a large proportion, as the research object, conducts partial discharge experiments on it, extracts partial discharge characteristic information, and proposes an improved Wasserstein generative adversarial network to train partial discharge spectra. Sample data, so as to generate more new spectral data, and then put the generated sample data and the original sample data into deep residual network training at the same time.