An online detection method for capacitor voltage transformer based on load classification

Abstract

capacitor voltage transformers (CVT) are widely used in the power system due to their good insulation and low cost. An accurate measurement performance provides critical information to ensure the safe and efficient operation of the power system. Therefore, online measurement error detection has received extensive attention. However, the problem of detecting CVTs deployed in substations with fluctuated loads has not been solved, because the frequent switching of these loads changes the size of monitoring indicators, resulting in misjudgment. In this paper, an online detection method based on load classification has been proposed. Several identification parameters are first put forward to classify loads. With these parameters, low-load modeling and monitoring data of the period of industrial users stably out of operation are screened out. Finally, online monitoring is achieved by the principal component analysis method. The efficacy of this method is verified in a 220kV substation actuating a steel plant.