Fast Probabilistic TTC Assessment Based on Copula Theory and Deep Ensemble Learning

Wind power integration makes the uncertainty of power system increasing dramatically, which brings new challenges to the calculation of Total Transfer Capability (TTC). A fast assessment method for probabilistic TTC based on Copula and deep ensemble learning is proposed. Considering wind power output uncertainties and correlation of wind speed in geographical close wind farms, Copula function is adopted to generate scenarios. Stacked Denoising Autoencoder (SDAE) is used to extract features directly from source data. Support Vector Regression (SVR) is selected as regressor and bagging ensemble strategy is applied to further improve the accuracy of the TTC estimation. The case study in simplified Shandong grid validates the effectiveness of the proposed method.