A feature extraction method of cable sub-cycle disturbance voltage

Abstract

When insulation defects arise in cables， they trigger the occurrence of sub-cycle disturbance voltage， making it imperative to extract the disturbance voltage features for timely defect detection and ensuring safe operation. To address this need， a method for extracting cable sub-cycle disturbance voltage features is introduced. Firstly， the cable’s sub-cycle disturbance voltage waveform is obtained through testing， and two characteristic parameters， namely time and amplitude， are defined for the disturbance waveform. Subsequently， voltage amplitude changes are rapidly tracked using the strong tracking Kalman filter （STKF）. Adaptive parameters are established to suppress fluctuations in the parameters of the Gaussian mixture model （GMM）， facilitating the extraction of disturbance voltage characteristic parameters through the combined use of STKF and GMM. The algorithm is then tested using disturbance voltage data obtained from experiments， and the effects of adaptive parameters， sampling frequency， and the detection algorithm’s dead zone within the Gaussian mixture model on the extraction of characteristic parameters are analyzed. MATLAB simulation results demonstrate that the computational outcomes remain relatively stable when the voltage sampling frequency is set at 2.25 kHz and above. This validation underscores the method’s capability to extract characteristic parameters related to cable sub-cycle disturbance voltage.

Keywords

• cable insulation defect
• sub-cycle voltage
• strong tracking kalman filter
• gaussian mixture model