Comprehensive evaluation of voltage sag severity level based on DBN

With the development of high and new technology, more and more new electrical equipment sensitive to voltage sag are being connected to the power grid. Therefore, it is urgent to improve the method for evaluating the severity of voltage sag. Based on the traditional voltage sag indexes, 8 types of voltage sag evaluation indexes were extracted in this paper, considering the three-phase imbalance, phase jump and other issues. The selection of these new indexes combined the measured waveforms of the grid side with the general voltage tolerance curves of the sensitive load of the user side. Based on the mutual information entropy, the thresholds of different severity of each index were determined, and the severity levels were divided. A method of evaluating the severity level of voltage sag based on DBN was proposed. The clustering ability of DBN was improved by training the model with measured data, and the attribution of severity level of each voltage sag event was determined to achieve the purpose of comprehensive evaluation. The updating of the DBN internal weights did not require human intervention, which reduced the excessive reliance on subjective experience in traditional methods. The effectiveness of this method was verified by a practical example.