A framework for automatically cleansing overvoltage data measured from transmission and distribution systems

Abstract Overvoltage data collected by the measurement systems in substations contain vital information about the power grid. With the development of smart grids, the mining and processing of this overvoltage data are of increasing significance. Unfortunately, there is currently no way of preventing error waveforms from being incorporated into the overvoltage database. In this paper, a cleansing framework for the measured overvoltage data is proposed. By automatically extracting and clustering features, error waveforms at the substation can be successfully isolated. In terms of implementation, a step-by-step scheme for pre-cleansing and full cleansing is established to improve the quality of the overvoltage data. Experimental results using measured overvoltage data demonstrate the effectiveness of the proposed solution. The proposed method and framework support the cleansing of overvoltage data, and are conducive to the analysis of the generation, propagation, and distribution characteristics of the overvoltages.