(Data alongside) Fault prediction of gas-insulated system with hypersensitive optical monitoring and spectral information
Description
Fig. 1. Phase-resolved scatterplots (φ, q), double-spectral scatterplots (Iband1, Iband2) for corona discharges occurring along the needle-plane gaps with different gap distances (d). (a)-(c): (φ, q); (d)-(f): (Iband1, Iband2); Fig. 2. Integral spectra of the corona discharge in SF6 (gap distance = 50 mm; gas pressure = 0.3 MPa; f = 15). A fiber-optic spectrometer (200-1150 nm), a collecting lens installed outside the quartz glass window of the test chamber, and a 19-core silica fiber were employed to measure the integral spectra of the discharges. Fig. 7. Light intensity stochastic distributions in an RGB 3-D coordinate system for the three types of defect. Fig. 11. Scatterplots of the DBSCAN clustering results for a double-sourced discharge (floating discharge and creeping discharge occur simultaneously with a discharge level of 320 pC). Fig. 15. Results of clustering analysis on TSS data (Cluster 1 and Cluster 2) and the evolution paths of the two types of discharge (Corona discharge: 30 μJ→93 μJ; Floating discharge: 77 μJ→224 μJ).
Files
Steps to reproduce
All the dataset is formatted in .txt.