Research on Data-Driven Transient Voltage Stability Assessment Considering Uncertainty

Abstract

In order to evaluate the transient stability of renewable energy grid more effectively and measuring the epistemic uncertainty of the evaluation results, a power system transient voltage stability assessment method based on temporal convolution stochastic differential equation network (SDE-TCN) is proposed. The network takes the time series of the basic physical quantity measurement data of the power system as the input. Temporal residual convolution block based stochastic differential equation network is employed to extract the time series features. Meanwhile, the proposed method introduces the focal loss function to eliminate the imbalance of the samples and improve the accuracy of transient stability assessment. The drift network for learning prediction of stability and the diffusion network for learning uncertainty measurement in SDE-TCN network are trained respectively. As a result, test results on IEEE 39 bus system show that this method can not only accurately evaluate the voltage transient stability of new energy power system, but also effectively quantify the uncertainty of stability evaluation.