On-line Evaluation Method of CVT Internal Insulation State Based on Support Vector Machine

Abstract

Capacitive voltage transformer (CVT) is a measurement device widely used in high-voltage power grids. Evaluating CVT's internal insulation online and replacing faulty device are critical for power grid's safety. However, the existing online evaluation methods mainly relaying on installing auxiliary sensors inside the device, which may change its internal electric field distribution and makes the device easy to be broken. In response to these challenges, an online evaluation method for the internal insulation state of CVT based on its measurements is proposed. This method explores the internal connection between the internal insulation state of CVT and its measurement error information, and then inverts the change of the internal insulation state of the CVT by monitoring the change of the measurement information of the CVT, and realizes the real-time online evaluation of the internal insulation state of the CVT based on the support vector machine. Experimental analysis shows that this method can accurately evaluate the internal insulation state of the CVT without the need for power outages and external sensors.