Estimation of Voltage Sag Frequency Based on the Multiple Characteristic Factors

The existing methods of estimating voltage sag frequency (VSF) do not adequately consider the influence of multiple line characteristic factors, such as pollution degree, lightning grade, wind zone grade, and bird damage grade, on short-circuit faults. This gap results in significant deviations between estimation results and actual conditions. In practice, voltage sags mainly result from instantaneous short-circuit faults in transmission lines, and these faults are simultaneously influenced by two key factors: 1) environmental characteristic factors, e.g., lightning grade and wind zone grade, and 2) inherent line characteristic factors, e.g., line length and line run time. In view of this gap, this paper proposes a novel method to estimate VSF by introducing the line fault models that incorporates multiple line characteristic factors to enhance the estimation accuracy of VSF. The proposed method involves three main steps: 1) calculate the failure probability of the line by using the random forest prediction algorithm, 2) obtain the distribution of line fault locations based on adaptive kernel density estimation algorithm (AKDE), and 3) estimate VSF with the fault point method. The efficacy and practical utility of the proposed method are rigorously validated via numerical simulations and field data from power grids.