Calculation of Dry and Wet Surface Flashover Progress of Insulators Based on Multidimensional Data Processing
摘要
Environmental humidity significantly affects the performance of topological insulators. This paper conducted surface flashover experiments on insulators under both dry and wet conditions. The experiments captured four types of signals. The paper systematically analyzed these data and compared the flashover characteristics in dry and wet environments. Finally, based on the optical features obtained from the experiments, a decision tree algorithm was employed to classify the types of flashovers (dry or wet). At the same time, a BP neural network was trained using electrical, acoustic and thermal signal features to calculate the flashover progress percentage.